Skip to main content

Full text of "Effects of buoyancy on gas jet diffusion flames"

See other formats


/v :■..'/ 

NASA Contractor Report 191109 ^ 3& H 



Effects of Buoyancy on Gas 
Jet Diffusion Flames 



M. Yousef Bahadori and Raymond B. Edelman 
Science Applications International Corporation 
Torrance, California 



April 1993 



(NASA-CR-191109) EFFFCTS OF N93-31031 

BUOYANCY ON GAS JET (DIFFUSION 

FLAMES Final Report (Science 

Applications International Corp.) Unclas 

364 p 

G3/29 0171948 



Prepared for 

Lewis Research Center 

Under Contract NAS3-22822 



NASA 

National Aeronautics and 
Space Administration 



tarifQF CONTENTS 



Section 



10 



11 



12 



INTRODUCTION AND BACKGROUND 

EFFECTS OF BUOYANCY ON GAS-JET DIFFUSION FLAMES; Science Requirements 

Document; R. B. Edelman and M. Y. Bahadori; January 1987 

EFFECTS OF BUOYANCY ON GAS-JET DIFFUSION FLAMES; Conceptual Design 

Review; R. B. Edelman and M. Y. Bahadori; June 1986 

__ p _ rTC; OF RUOYANCY ON GAS-JET DIFFUSION FLAMES: EXPERIMENT AND 
W^.KBm^Lu. ?. Bahadori; mti Astrcnautica. Vol. 13. No. 11/12, pp. 
681-688,1986 

26th Aerospace Sciences Meeting, Reno, Nevada, January 1 988 

ri , |FT n.FFUSION FLAMES UNDER REDUCED-GRAVITY CONDITIONS, 

Institute, Houston, Texas, April 1988 

nyYPFN CONCENTRATION EFFECTS ON MICROGRAVITY LAMINAR METHANE AND 

October/November 1989 

MEASUREMENT OF TEMPERATURE IN MICROGRAVITY LAMINAR DIFFUSION 

December 1990 

IGNITION AND BEHAVIOR OF LAMINAR GAS-JET DIFFUSION FLAMESIN 
MICROGRAVITY; M. Y. Bahadori, R. B. Edelman, D. P. Stocker, and S.L. Olson, AJAA 
jQUIDal, Vol. 28, NO. 2, pp. 236-244, 1990 
am ANAIYTICAL SOLUTION FOR TRANSIENT, CYLINDRICALLY SYMMETRIC 

Ohio, May 1990 

Aerospace Sciences Meeting, Reno. Nevada, January 1990 

MAA 9?0719 MM 29th Aerospace Sciences Meeting. Reno, Nevada. January 1991 



TABLE OF CONTENTS (p 9 nt ) 



Section 



13 EFFECTS OF OXYGEN CONCENTRATION ON RADIATIVE LOSS FROM wnpuAi 

?.".? £ d ! lman ' D - P - s,ocker - R - G- Sotos, and D. F. Vaugharv Paper AIAA-92-0243 
AIAA 30th Aerospace Sciences Meeting, Reno, Nevada, January 1992 

14 F. R A A JiQ A !! u AL H INFI -. U S N S E f ON THE BEH AVIOR OF CONFINED DIFFUSION 
FLAMES; U. Hegde and M. Y. Bahadori; Paper AIAA-92-0334, AIAA 30th Aerosoace 
Sciences Meeting, Reno, Nevada, January 1992 Aerospace 

15 COMBUSTION OF GASEOUS FUELS UNDER REDUCED-GRAVITY CONDITIONS 
M. Y. Bahadon and R. B. Edelman; Paper LB-038, in press in Proceeding nf th„ a^M 
SYmPQSlum on Lunar BafiftS, Lunar and Planetary Institute, Houston, Texas, 1992 

16 EFFECTS OF OXYGEN CONCENTRATION ON RADIATION FROM MICROGRAVITY 
LAMINAR PROPANE DIFFUSION FLAKES; M. Y. Bahadori £ B Ke.r^n. R° G ^tos 
D. P. Stacker, and D.F. Vaughan; 1992 ' 

17 PRELIMINARY OBSERVATIONS ON THE EFFECTS OF BUOYANCY ON 
TRANSITIONAL AND TURBULENT DIFFUSION FLAMES; M Y B?had?ri 
S' L XW an :P- ? Stocker ' K - J - Weiland, and R. B. Edelman; Paper presented at the 
Central States Meeting of The Combustion Institute, Columbus, Ohio, April 1992 

18 EFFECTS OF BUOYANCY ON LAMINAR GAS-JET DIFFUSION FLAMES Science 
Requirements Document (Space Experiment); M. Y. Bahadori and R. B. Edelman; October 

19 EFFECTS OF BUOYANCY ON GAS-JET DIFFUSION FLAMES; Concept Design (Space 
Experiment); M. Y. Bahadori and R. B. Edelman; December 1991 9 ( P 



11 



SfifitlflDJ. 



Introduction and Background 



Introduction and Background 



This final report documents the research findings of Contract NAS3-22822 
performed by Science Applications International Corporation, entitled "Effects of 
Buoyancy on Gas Jet Diffusion Flames". The work was performed in two 
phases, as follows: 

(a) Phase I, from October 1981 through October 1987; this phase 
consisted of a modest effort involving only a fundamental investigation of 
combustion in gas jet diffusion flames, a feasibility study for a space 
experiment and preparation of the associated Science Requirements 
Document (SRD) and Conceptual Design Review (CoDR); 

(b) Phase II, from October 1987 through April 1992; this phase consisted 
of a comprehensive effort involving experiments in the 2.2-Second Drop 
Tower and the 5.18-Second Zero-Gravity Facility of NASA Lewis 
Research center, and preliminary tests in the KC-135 research aircraft, in 
addition to the associated theoretical modeling and data analyses. 



The principal objective of the effort was to gain a better understanding of the 
fundamental phenomena involved in laminar gas jet diffusion flames in the 
absence of buoyancy by studying the transient phenomena of ignition and flame 
development, steady-state flame characteristics, soot effects, radiation, and , if 
any, extinction phenomena. This involved measurements of the flame size and 
development, as well as temperature, radiation, and (the originally planned) 
species concentration. Additionally, flame behavior, color, and luminosity were 
observed and recorded. The test quantified the effects of Reynolds number, 
nozzle size, fuel reactivity and type, oxygen concentration, pressure, and jet 
momentum on flame characteristics. Analytical and numerical modeling efforts 
were also performed. 

During Phase I (in 1986) a space experiment was proposed and presented to 
NASA Headquarters at a Conceptual Design Review (CoDR). In 1 987, NASA 



Headquarters recommended that a comprehensive series of ground-based tests 
in the drop towers and tests using a research aircraft be conducted prior to a 
second CoDR. This led to the initiation of the second phase of the effort. In 
December 1991, the Phase II results were presented at a CoDR. NASA 
Headquarters recommended that a space experiment be re-proposed following 
the 1993 NASA Research Announcement (NRA). 

This final report contains the 1986/87 Science Requirements and Conceptual 
Design documents, journal publications and meetings presentations resulting 
from the second phase of the effort, and the 1991 SRD/CoDR documents. It 
should be noted that the data and associated modeling results (only a 
representative set is presented in this document) are currently in the final phase 
of compilation, reduction and documentation. This will result in a substantial 
number of publications (a few preliminary versions of which are presented in this 
report) to be submitted to archival journals during the remaining months of 1992. 
This effort will be carried out while a new effort (Contract NAS3-25982), entitled 
"Effects of Buoyancy on Transitional and Turbulent Gas Jet Diffusion Flames" is 
underway. This new effort was initiated in November 1991. Due to the overlap 
between the two programs, the last meeting presentation of this report contains 
some preliminary results from the new effort. 

The remainder of this section contains an overview of the major findings of the 
2.2-second tests, the 5.18-second tests, the (preliminary) KC-135 tests, and the 
theoretical modeling effort. 

la) The 2.2-Second Drop-Tower Tests 

These tests were conducted in order to characterize the behavior of flames, and 
to define the matrix for the 5.18-Second tests. Methane and propane flames 
were tested in quiescent oxidizing environments using different nozzle sizes 
(0.051 to 0.0825 cm in radius) and fuel flow rates (0.3 to 1.5 cc/sec for propane 
and 1 .0 to 5.0 cc/sec for methane). In these experiments, the Reynolds number 
was varied from approximately 20 to 200, pressure from 0.5 to 1.5 atm, and 
oxygen volume fraction from 18% to 50% (using nitrogen as diluent). The flames 
were also ignited in microgravity, which demonstrated for the first time the ability 
to ignite and sustain a gas jet diffusion flame in a low-gravity environment. 



li 



The results showed that the 2.2 seconds of microgravity testing was not sufficient 
for the flame development to approach its near-steady state behavior. In 
addition propane flames in high-oxygen-concentration/high-pressure 
environments showed significant soot formation resutting in flame-tip opening 
with a column of soot emerging from the tip of the flame. Low-pressure/low- 
oxygen-concentration flames were found to be extremely weak, blue in color with 
no visible soot emission, and had an open tip. The results also showed that 
microgravity flames are taller, wider, more diffuse, and have no flicker compared 
,o the corresponding normal-gravity flames. These microgravity flames take 
much longer time to reach a near-steady state. The tests also quantified the 
effects of pressure, oxygen concentration, nozzle size, Reynolds number, and 
fuel type on the characteristics of microgravity hydrocarbon diffusion flames. 

The observations and associated measurements on global flame characteristics 
provided a database for the 2.2-second microgravity flames, and were used to 
develop the test matrix for the longer-duration microgravity expenments 
described below. 

( h) The 5 1 q-f^nnd MifTW auit '' tests 

These tests were conducted to study the behavior of flames under longer- 
nation microgravity conditions, and to quantify both flame »"^|™ «£ 
,,ame radiation. Global flame radiation was measured us.ng a th rmopHe 
detector, which was the first demonstration of the ability to measure radiat on 
from a flame in a microgravity environment. Temperatures were measured 
simultaneously a. nine different locations (using a 3x3 rake of thermocouples) 
Zu provided time-resolved temperature data both inside anc I outside .he 
name The results showed significant effects of microgravity on radiatrve losses 
and emirature Md outside the flame, in addition to the globa, characters of 
he flam? such as tip-opening in propane „ames, and «™ «•■« "£ 
luminosity for the 5-second period. In addition, it was shown "ough he 
flames appear to reach their near-steady state behav,or towards the end of Mft. 
L quantitative measurements point to a transient behavior, and that flames 
Z to be tested a, times longer than 5 seconds to fulry quantify the,r steady- 
state characteristics. 



ill 



Methane and propane flames were studied with one nozzle size (0 0825 cm in 
radius), with flow rates of 1.0 and 3.0 cc/sec for methane and 0.5 and 1 5 cc/sec 
for propane (when converted to the standard conditions). Oxygen concentration 
in a nitrogen diluent was also varied from 18% to 30%, while the environmental 
pressure was varied from 0.5 to 1 .5 atm. 

The results showed, for example, that flame far field experiences noticeable 
temperature rise in the absence of buoyancy; this rise in temperature dose not 
reach its steady state value at the end of the drop. In addition, continuous 
change in the color and luminosity of the flame was observed due to the transient 
response of the field surrounding the flame. Also, radiation measurements 
showed that microgravity flames can have radiative losses of up to an order of 
magnitude higher than their normal-gravity counterparts. This has a significant 
implication on radiative losses and radiative ignition in a microgravity 
environment. Two flames were studied at 1 atm (for propane) and 0.5 atm (for 
methane) with 15% oxygen concentration in a nitrogen environment These 
flames were found to be entirely blue, extremely diffuse, and open-tipped Yet 
the rad.at.on levels were quiet significant compared to those in normal-gravity 
Th.s indicates that although soot formation and subsequent emission were 
reduced, contributions from the gas-phase species were quiet significant due to 
the absence of buoyancy; radiation, although lower, was of the same order of 
magnitude in blue flames compared to the luminous flames in microgravity. 

The test matrix was not completed due to the closure of the Zero-Gravity Facility 
for renovation. However, a major part of the matrix was completed (except the 
intermediate flow rates of 2 cc/sec for methane and 1 cc/sec for propane) The 
ongmal plan was to conduct gas-sampling measurements at mirror-image 
locations of the thermocouples with respect to the flame centerline. The 
expenmental chamber had provisions for performing this task. However, this 
effort was not conducted due to the shut-down of the facility. 

A substantial amount of data on flame size, color, tip characteristics 
development, radiation, and temperature as a function of fuel type, Reynolds 
number, pressure, and oxygen concentration were obtained. Representative 
data are presented in different parts of this report. Data reduction and associated 



IV 



documentation are currently in progress, and the results will be published in the 
open literature, in the near future. 

(r) ThftKC - 135 Tests 

These preliminary flight tests were conducted to study the effects of longer 
durations of low-gravity environment on flame behavior. The intent was to 
conduct measurements to quantify flame radiation, temperature, spec.es, and 
global characteristics, and to determine the effects of acceleration and g-g.tter. 
Tests were conducted for methane and propane flames with one nozzle size. 
Observations on flame behavior were not conclusive, but showed that the overall 
characteristics of flames can be obtained for the attached (1 0*g) and free-float 
(10-3g) configurations of the test package, that flames are strongly influenced by 
the g-level and g-jitter, and that an average time in the order of 5 seconds m the 
free-float case and 10 seconds in the bolted-down case may be obta.ned. 
although g-jitter affects the flame. No quantitative measurements were 
performed in these tests. 

( d) Theq rfttirr al Modeling 

During the course of the second phase of the program, various analytical models 
were developed. These included a transient model for the laminar diffus.on 
flames a model for the transient filling process in an evacuated or partially 
evacuated gas sampling bottle, and a model for the effects of g-jitter and partial 
gravity on the flame response. 

A comprehensive numerical model which incorporates the effects of gravity, 
diffusion, inertia, viscosity, and combustion was modified to include gas-phase 
and soot radiation and multi-component diffusion. Work was also completed on 
the formulation of soot processes, thermophores*, and finite-rate quaa-globa 
kinetics. These submodels were not incorporated into the model by the end o 
the contract. The developed model is two-dimensional (cylindrical or rectangular) 
and solves the conservation equations for mass, momentum, energy, and 
species in their boundary-layer form. An explicit finite-difference scheme ,s used 
to solve these governing equations, using the von Mises transformation. 



Selected predictions from the analytical and numerical models are presented in 
various sections of this report, and the findings are currently in the process of 
being submitted for journal publications. 

The following sections present the SRD/CoDR documents, and associated 
publications and meeting presentations of the entire effort in a chronological 
order. 



VI 



S§£ii£Q_2 

'Effects of Buoyancy on Gas-Jet Diffusion Flames' 
- Science Requirements Document- 



R. B. Edelman and M. Y. Bahadori 



January 1987 



i — Science Applications International Corporation- 



EFFECTS OF BUOYANCY ON 
GAS JET DIFFUSION FLAMES 

- Science Requirements Document - 

Submitted by: 

Raymond B. Edelman (Principal Investigator) 
M. Yousef Bahadori (Assistant Investigator) 

SCIENCE APPLICATIONS INTERNATIONAL CORPORATION (SAIC) 

Combustion Science and Advanced Technology Department 
9760 Owensmouth Avenue 
Chatsworth, California 91311 



Submitted to: 
NASA-LEWIS RESEARCH CENTER 
21000 Brookpark Road 
Cleveland, OH 44135 



January 1987 




TABLE OF CONTENTS 

Page 

I . INTRODUCTION 1 

II. BAC KGROUND 1 

III. OBJECTI VES 19 

IV. CURRENT MODELING STATUS 21 

1. Steady Flow Jet Diffusion Flame Analysis 21 

2. Analytical Solution for Transient Jet 

Diffusion Flames 26 

3. Transient, Navier-Stokes Based Model 33 

V. DATA NECESSARY TO VALIDATE/ IMPROVE MODEL(S) 35 

VI. IDENTIFICATION OF EXPERIMENTS 37 

VII. APPARATUS AND CONDITIONS 42 

VIII. DATA ANALYSIS 42 

References 49 



1 introduction 
Although diffusion flames have been the subject of research for 
rany hey are hot well understood and, while the effects of gravny 

s1ble for these changes have yet to be determined. The goal of 

h to develop an improved understanding of gas oet di fusion 
flame s The laminar gas jet diffusion flame was selected for study 
[I:! lt embodies mechanisms important in .— - 
as occur in fires and practical combusts system . Furt ™ re 
t„ its relative simplicity and ease of control, it lends itself to 
r t ca, end experimental studies. However, existing t eore- 
a models are deficient and improved models are reared becaus 
a essential to the interpretation of data, providing a mean of 
h e re.ically isolating mechanisms influencing the o serve a r 
The complexity of these mechanisms and their coupling establish he 
, re for data with which to verify assumptions inherent ,n the 
gu t ,„<« If cavity is eliminated, then its effect can 

A*t* i* <;pverelv limited and incomplete. The exisnny 

ril rutltive visual observations on ^^^^ 

,„ a state of transient adjustment at the end of the 2.2 sec. drop 

tower test time. 

This doc^ent describes the requirements of an «■»''«"* "* ^ 
, I in other low-gravity environments, would help to eliminate 
performed in otner low yiaviur .. ff .:.- flames, 

current deficiencies in our understanding of gas jet dlftusion 

tt. Background 
The problem of fire safety has received the attentio, -of research- 
ers for many years which has resulted in a continuous effort to 



-es 
?ss 

ic 



stand the complicated processes occuring in fi res . Recently, interest 
m the fire problem has been expanded to include aerospace applications 
due to the concern over fires aboard spacecraft. The primarv conclusion 
drawn from the existing base of information on fires is that a more 
fundamental understanding of diffusion flames is needed in order to pre 
diet their behavior and to control them. Although the effects of gravity 
on a specific burning process have been observed, the mechanisms respon- 
sible for these changes have yet to be determined. A critical limitation 
of the existing earthbound experimental data obtained in the 2 2-sec 
drop tower is that only limited elapsed time is available to fully charac- 
tenze the flame behavior under reduced-gravity conditions. 

Diffusion flames involve convective and diffusive effects and fir 
are a complicated type of this class of combustion process. Reqardle 
of the type of fuel involved (i.e., liquid, solid or gaseous), the basi. 
mechanisms which control the fires are the same. These include the proc 
esse, of mixing, radiation, chemical kinetics, soot formation, diffusion 
and buoyancy-induced convection. The laminar gas jet diffusion flame 
embodies these mechanisms while representing a basic element of more 
complex, convectively induced turbulent flames. 

In order to understand the mechanisms controlling a fire, labora- 
tory-scale turbulent gas jet diffusion flames have been extensively 
studied. These small-scale flames provide important information in 
relation to the processes occuring in practical combustion syslems as 
well. However, turbulent flames are not fully understood and their 
understanding would benefit from more fundamental studies of laminar 
diffusion flames in which, under normal-g conditions, the complication 
of convection-induced turbulence is eliminated and the interplay of 
transport phenomena and chemical kinetics becomes more tractable. But 
even this basic flame is not completely characterized in relation to 
soot formation, radiation, diffusion, and kinetics. Therefore, gain- 
ing an understanding of laminar flames is essential to the understand- 
ing of fires ir, which the same basic pheonomena occur. By working 
under microgravity conditions, the effects of buoyancy can be isolated 
and the results would form a basis to address, in logical order, the 
fundamental processes occuring in diffusion flames (see Edelman'and 
Bahadori, 1986). 



In the remainder of this section, we describe a cross-section of 
past experimental and theoretical efforts in relation to laminar dif- 
fusion flames under both normal-gravity and reduced-g conditions. 

Laminar gas jet diffusion flames (Fig. 1) have been a subject of 
extensive analytical, numerical and experimental research since the 
appearance of the classical work of Burke and Schumann (1928). Most 
of the analytical studies to date deal with the mathematical represen- 
tations of flame shapes based on the Burke-Schumann methodology which 
neglects almost all of the phenomena critical to understanding the 
structure of these flames. These include transient effects, soot 
formation, non-uniform flow parameters, variable transport properties, 
coupling between the conservation equations, and finite-rate chemical 
kinetics in all of the analytical approaches, the effect of buoyancy 
had to be neglected (i.e., high-momentum flows) in order to enable one 
to obtain a closed-form mathematical representation of flame shape, 
mass-fraction distribution and temperature distribution. In addition, 
the assumptions cited above render the theory of little value in under- 
standing the structure of diffusion flames. Applications of this theory 
to laboratory-scale gas jet diffusion flames (Burke and Schumann, 1928; 
Astavin and Ryazantsev, 1979; Penner, Bahadori and Kennedy, 1984) and to 
the flamelets of composite solid propellants (Beckstead, 1981) have shown 
only qualitative agreement between the theoretical and experimental 
flame shapes. Table 1(a) describes a selected number of analytical 
works performed on laminar gas jet diffusion flames. Few numerical 
investigations of these types of flames under normal-g and microgravity 
conditions have been attempted. The results of numerical calculations 
of the set of conservation equations for normal-g (Mitchell, Sarofim 
and Clomburg, 1980) and zero-g flames (Edelman et .1 . 1973) have shown 
that improved understanding is needed in relation to the mechanisms 
occuring under these conditions. The mechanisms include "diation, 
soot formation, chemical kinetics and transient phenomena. Table 1(b) 
shows a selected number of numerical modeling approaches. Expen- 
me ntal investigations of nor-mal-g flames have provided a large amount 
of information which is not obtainable from the theoretical studies 
cited above, i.e., flame structure, soot mass fraction and number 
density, and temperature, species and velocity distributions (see, 
for example, Mitchell, Sarofim and Clomburg, 1980, and Flower and 




Fig. 1. Schematic diagram of a confined 1 
flame. 



aminar gas jet diffusion 



Table 



1(a). Selected Analytical Studies on Laminar Gas Jet Diffusion Flames. 



AUTHOR(s) 



Burke and 

Schumann 

(1928) 



Astavin and 

Ryazantsev 

(1979) 



Penner, 
Bahadori and 
Kennedy (1984) 



KETHCD 



CHARACTERISTICS 



Analytical represent- 
ations of the shapes of 
laminar, cyl indr ical ly- 
symmetric, co-axial gas- 
jet dif'usion flames 
under steady-state 
condi tions . 



Assumptions of equal and 
constant velocities , 
transport coefficients, 
and physical properties 
of the two gas streams; 
flame-sheet approxi- 
mation; Lewis number of 
unity (which decouples 
the energy and species- 
conservation equations), 
constant pressure (which 
el iminates the momentum 
equation) ; absence of 
buoyancy; no axial 
di ffusion. 



DATA USED 



Beckstead 
(1981) 



Edelman and 
Bahadori (see 
the present 
document) 



Same as Burke and 
Schmann (1928) but for 
fuel and oxidizer sep- 
arated by a flow of an 
inert gas (a) from each 
other and (b) from the 
wall for both cylindrical 
and rectangular burners 



Same as Burke and 
Schumann (1923) . 



Same as Burke and 
Schumann (1928) . 



Hethane, city gas, 
and CC flames are 
used. Air flow 
rate*6- 4 2 ftVhr. 
fuel flow rate* 
2-25 ft 3 /hr. 
Inside radius" 
1/4-5/16 in. 
Outside radius" 
1/2-5/8 in. 



DEFICIENCIES 



No acco 
format 
ducts o 
i on; 1 ac 
fini te 
etics a 
format i 
anism; 
of radi 
axial d 
assumpt 
coupled 
ion equ 
transie 



urt of the 
on of pro- 
f combust- 
k of 

rate kin- 
nd soot- 
on mech- 
no effects 
ation; no 
iffusion; 
ion of de- 
conservat 
ations; no 
nt effects 



CCWEN T S 



Theoretical 
f 1 ame hei g K ts 
irt compared 
with the ex- 
perimental 
resul ts 
under normal - 
g ; qua 1 i ta- 
tive agreemer 
is obtained 
between ex- 
periment and 
theory. 



Analytical representation 
of the heights of 
laminar, cyl indrically- 
symnetric flamelets of 
composite solid propel - 
lants under steady-state 
condi tions . 



Same as 8urke and 
Schumann (1928) but for 
constant arbitrary flow 
velocities, transport 
parameters and physical 
properties in the pre- 
sence of axial di ffusion 
with unimolecular decom- 
position reactions of 
fuel and oxidizer. 



Same as Burke and 
Schumann (1928) but in 
the presence of axial 
di ffusion . 



Analytical representation 
of the shapes of laminar 
cyl indrically-symnetric, 
transient, co-axial gas- 
jet diffusion flames. 



Same as Burke and 
Schumann (1928) but ir 
the presence of axial 
diffusion. 



Parametric values 
of velocities , 
di ffusion coef- 
ficients and tube 
diameters are used 
for methane/air 
flames . 



Same as Burke and 
Schumann (1928) . 



The data of Burke 
and Schumann (1928) 
and Mitchell et al 
(1980) are used. 



Fami 1 les of 
flame shapes 
and temper- 
ature contour 
art obtained . 
'•0 comparison 
wi th data . 



Same as Burke and 
Schumann (1928) 
exceot for the 
effect of axial 

dl ffusion. 



T ypical coeffic- 
ients of df ffusion 
and velocities for 
evaporating solid 
propel lant species 
are used. 



The results 
are in qual- 
itative agree 
ment with ex- 
perimental 
data . 



Same as Burke 
and Schumann 
(1928) except for 
the effects of 

axial diffusion. 



The data of Cochrar Same as Burke and 
(1972) and Cochran Schumann (1928) 
and r-asica (1970) except for the 



Flame stand- 
off distances 
are obtained . 
No data avai 1 
able. 



are used, e.g. , 
nozzle radius'0.186 
cm, methane velo- 
city^. 8 cm/sec, 
etc. 



effects of axial 

diffusion and 
transient pheno- 
mena . 



Flame devel- 
opment wi th 
time is ob- 
tained. The 
results in- 
dicate that 
times greater 
than 2.2 sec 
are needed for 
flame to 
reach steady 
state under 
zero-g 
conditions . 



Table 1(b). Selected Numerical Studies on Laminar Gas Jet Diffusion Flames 



AUTHOR(s) 



rtE'HCC 



Mitchell , 
Sarofim and 
Clomburg 
(1980) 



Numerical solution of 
governing conservation 
equations for mass, 
momentum, species, and 
energy, using a global 
chemical reaction at 
the flame boundary with 
infinite rate. A steady- 
statt normal-g methane/ 
air flame was studied 
which was laminar, 
confined, and axisy- 
mmetric . 



CHARACTERISTICS 



Flame-sheet concept for 
locating the stoichio- 
metric fuel -oxygen 
interface and hence, the 
points of heat release. 
Natural convection 
effects and variable 
thermodynamic and trans- 
port properties are 
allowed. Both axial 
and radial convection 
and di f fusion are in- 
cluded. 



DATA USED 



Klajn and 
Oppenhetm 
(1982) 



Inner tube dia- 
meter. 1 .27 cm; 
outer tube dia- 
meter^. 18 cm, 
methane flow rate' 
5.7 cm-/sec; air 
flow rate-187.7 
cm 3 /sec. 



DEFICIENCIES 



No detailed 
chemical kin- 
etics, soot 
formation or 
radiation. 



The numerical model 
treats the effects of 
exothermici ty on the 
contours of gaseous , 
unconfined jet diffusion 
flames with negligible 
buoyancy effects. 



Edelman, 
Fortune, 
Wei lerstein, 
Cochran, and 
Haggard (1973 



Numerical solution of 
conservation equations. 
It is shown that kinetic 
effects are of primary 
importance in flames 
under reduced gravity. 
Other controlling fact- 
ors are axial diffusion, 
radiation and transient 
effects. 



Simplifying assumptions 
of flame-sheet approxi- 
mation, Schmidt and 
Prandtl numbers of unity 
and constant specific 
heats. Solutions are ob- 
tained in closed alge- 
braic form. Axial 
diffusion is not included 



Nondimensional ued 
values of diameter;, 
stoichiometric 
ratios, velocities , 
etc. are used in 
a parametric form. 



Lewis and Prandtl 
numbers of unity; 
infi ni te-rate 
global chemical 
reaction; no rad- 
iation or soot 
forma t i on . 



The model includes the 
coupled effects of 
inertia, viscosity, 
diffusion, gravity, and 
combustion in a boundary. 
layer type formulation. 
Axial and radial con- 
vection and axial dif- 
fusion are considered. 



Comparison wi th 
drop-tower data of 
Cochran and Masica 
(1970) and Cochran 



:i972 
Table 1 



see 

c). 



No detai led kin- 
etics, soot form- 
ation, axial di f- 
fusion, and rad- 
iation. 



CCf ,u E'. 



e^oera turps 
spec es con- 
centrations 
and ve' oc i t- 
ies are ob- 
tained "tie 
resul ts shea- 
the import an 1 
effect of 
buoyancy, by 
which, the 
velocity in- 
creases by a 
factor of AC 
inside an 
overventi la t- 
ed CH^-air 
flame. Good 
compa r i son 
with e« per i - 
mental data 
is obtained 



r air agree- 
ment when 
coirpared >< t 
the experi- 
mental data 
under zero-^ 
Condi t i ens 
of Cochran 
(197?) and 
Haggard and 
Cochran 
(19-3) 



Good agree- 
ment for 
norma I - g 
flames and 
fair agree- 
ment for 
steady-state 
low-Re. zero- 
g flames of 
NASA-Lewis 
drop-tower 
experiments . 



ORIGINAL PAGE IS 
OF POOR QUALITY 



Bowman 1983). Efforts to isolate the buoyancy effects in laminar 
les'hal I ^r been lifted to the 2.2-sec. NASA-Lewis Research 

Center drop-tower (gravity levels- 10" 5 ,) experiments (Cochran an 

sica 1970; Cochran, 1972, Haggard and Cochran, 1973; Haggard, 1981) 

• n, m ,rical modeling (Edelman, Fortune and Weilerstein, 
and accompanying numerical mooenng \<- 

,973- Edelman et al . 1973). The drop-tower experiments comprise of a 

self-contained gas jet diffusion flame apparatus which is allowed o 

fall inside a falling drag shield. Once establishing an 

the steady-state normal-g flame, the apparatus is dropped and the fl m 

o s rv d throughout the duration of the f.ll . providing t-e-res 

J on the development of the flame shape for the near-zero, , co tion 

achieved during the drop. Attempts have also been made to study the 

9 jet diffusion flames under elevated gravity conditions achiev n 

ce trifuge (Altenxirch et al. 1976) and under negative-g conditions 

; nverting the flame using a downward-flow configuratio, , £„r, ■ * 

, ,, a TQRi- SAIC's recent inverted flame data). Studies 
utawa 1961; Haqqard, l9ol, i>Mit b Iac,a 

be c ndu ted on soot Nation from laminar diffusion flames under 

r ie 8 'p -re on soot formation in the absence of buoyanc, 

us of selected experimental wor k s on laminar flames are prese t m 
~. e 1(0. We see that while the effects of buoyancy are apparent in 

of [he studies presented in Tables ,(a,-UC, it has not bee, is net- 
, y isolated (except for the flames of drop-tower experiments), and 
large gap in the data base exists in this context. 

On the basis of the research cited above, we have identified those 
mechanic ^diffusion flames which are in need of further uners a - 
ing from both theoretical and experimental points of ,« h.. .-eh 
anisms include chemical kinetics, radiation, transient effects, soot 
formation, and axial diffusion. 

Kinetic effects are probably the most complicated phenomena when 
combi ith soot generation. Hydrocarbon diffusion flames show sig- 

flcant amounts of soot nation and agglomeration un e, - no™ 1- 
conditions and atmospheric pressure. Increase in soot yield h n 
observed (Flower and Bo-an, 1983) under elevated pressures. In the 



Table 1(c). Selected Experimental Studies on Laminar Ga 



s Jet Diffusion Flames 



AUTHOn(s) 



Mitchell , 
Sarof im 
and Clomburg 
(1980) 



METHOD 



Confined, axisynmetric , 
laminar methane-air 
diffusion flames. 



CHARACTERISTICS 



DATA USED 



Distributions of temper- 
ature, velocity and 
species were measured. 
The effects of buoyancy 
have not been isolated. 



Inner tube dia .» 
1 .27 cm; outer 
tube dia .'5.18 cm, 
methane flow rate- 
5.7 cm3/sec, air 
flow rate=lR7.7 
cmVsec . 



DEFICIENCIES 



CCVVEVS 



Flower and 
Bowman (1983) 



Structures- of 2-0, lam- 
inar ethylene-air flames 
under normal -g and 
elevated-pressure cond- 
itions are studied. 



No kinetics, SOOt 

formation or rad- 
iation effects . 



Soot mass fraction and 
number density is ob- 
tained. No significant 
effect of pressure on 
the size of soot parti- 
cles is observed. 



Pressure range" 
1.0 - J. 5 atm. 
A rectangular 
burner was used 
with air velocity 
■ 22 cm/sec and 
fuel velocity 
7 cm/sec. 



Only gas temper- 
ature and soot 
data are obtained 



Cochran and 
Masica (1970) 



NASA/Lewis drop-tower 
experiments (see the 
text for detai's) . 



Cochran 
(1972) 



Flame behavior is ob- 
served during the 2.2- 
sec drop of the test 
chamber. Quiescent air 
for all of the drop- 
tower experiments ex- 
cept those of Haggard 
(1981) which were stud- 
ied for inverted flames 
(under normal -g cond- 
itions) and zero-g 
flames in the presence 
of forced-air velocities. 
see the text for details. 



CHaVair flames; 
nozzle radius' 
0.186 - 0.442 cm; 
methane flow rate* 
1.2 - 5.3 cm3/ sec . 



Only flame 
lengths and be- 
haviors are ob- 
served. No 
measurements on 
quantitative 
data was possibli 
due to the short 
periods of drop. 



CHa/air flames; 
nozzle radius- 
0.051 - 0.113 
cm; fuel flow 
rate-0.7 - 12.2 
cm3/sec. 



Importance of 
buoyancy is 
demonstrated , 
by which, the 

ve'octy in. 

c-e»<ps by a 
'actor of 40 
ms ide an 
overvent 1 1 a te^ 

CH«-air flame 

The data are 
in good agree 

ment wi th the 

results of 

model i ng . 



Increase in 
soot yield 
'mass of 
soot/mass of 
fuel) as the 
pressure to 
t*ie power 
?.5 - 1.0 
was observed 

*Ai >fi (j m and 

i ntegrated 

SOOt n',T( 

fract icrs 
increased 
as the ores- 
sure to a 
power be- 
tween 1.5- 
2.D. 



Sudden incre- 
ase in gray i t 
level result- 
ed in ifTmed 1 - 
ate reduction 
n flame len- 
gth followed 
by flame ex- 
panding away 
from the 
burner and 
final ex ting - 
u i shment . 



Steady-state . 

transient and 
ex t inguished 
flames exist- 
ed in zero 
gravity. 
Flames were 
" SOX long- 
er and wider 
in zero-g 
than in 
normal -g . 



ORIGINAL PAGE IS 
OF POOR QUALITY 



Table 1(c) . Continued, 



author's) 



Haggard and 

Cochran 

(1972) 



Haggard 
(1981) 



METHOD 



CHARACTERISTICS 



Al tenkr ich , 
Eichhorn, Hsu, 
8ranc ic , and 
Cerallos 
(1976) 



Gas-jet diffusion flames 
under elevated gravity 
conditions achieved in a 
centri fuge. 



Hydrogen, methane, 
ethane and propane are 
used . 



Kimura and 
Ukawa (1961) 



Gas-jet diffusion flames 
under negat1ve-g condi- 
tions by inverting the 
flame using a downward- 
flow configuration. 



Data are obtained for 
both normal -g and in- 
verted flames. 



DATA USED 



Ethylene and pro- 
pylene d i f fus i on 
flames; nozzle 
radiuS'0.051 - 
0.083 cm; fuel flow 
rate'O.a - 3.6 
cm 3 /sec. 



CHj/air flames; 
fuel-nozzle radius 
0.05 - . 30 cm; 
coaxial , convergent 
am-nozzle radius = 
1.4 cm at exit 
Diane, fuel flow 
rate-1.55 - 10.3 
cmVsec; air flow 
rate=0-597 cmVsec 



DEFICIENCIES 



1 ,83-m diame'er 
centrifuge; burner 
inside diameter* 
0.05 - 0.21 cm; 
rotational speed" 
200 rpm. 



Only photographs 
are obtained and 
flame lengths an 
measured . 



C0'~ENT5 



ria^e length 
is described 
in ten-s of 
flow pa^a- 
ir.eters and 
fuel prop- 
erties . 
Stable zero- 
g lengths 
were observe 



Expenmcta I 
and computed 
flame lengths 
for nonra'-g 
have shown 
good agree- 
ment. Elar-e 
ex ti nguish- 
ment upon 
entry into 
zero-g »as 
studi ed . 
e ' a t iv e 1 y 
low force-- 
i t ve'oc ' t- 
•es (-\Cc-n' 
sec ', «fe 
suf r ' c ' e r t 
to susta'r 
como jS t i on 
in ze r o 
gravity. 



City gas flames 
are used. Fuel 
nozzle df a . -1 .5 . 3 
and 6 mm ; air duct 
dia . '52 and 96 mm; 
air flow rate - 
10 m/sec, fuel flow 
rate <. 40 m/sec. 



Only flame 
lengths are ob- 
served. 



The 
i nd 
tha 
Ten 
car 
i no 
ere 
1nc 
buo 
Als 
lif 
and 
uis 
wer 
ved 

9-1 
inc 



resu 1 ts 
icate 
t flame 
ath and 
bon lur- 
s i t y dB- 
ase with 
reas i ng 
yancy. 
o, flane 
t-off 

exting- 
hment 
e obser- 

as the 
evel 
reased . 



Flame length 
1s proport- 
ional to the 
fuel flow 
rate and in- 
versely pro- 
portional to 
the diffus- 
ion coeff- 
icient. 



*?5£38££ 



Table 1(c). Continued. 



author; s ) 



SAlC's 

inverted 

flame 



METHOD 



CHARAC r FRIS*ICS 



DATA USED 



A downward-flow configur- 
ation is used to study 
coaxial gas-jet diffusior 
flames . 



Temperature, flame shape, 
etc . are obtained. Stag- 
nation points are pre- 
dicted indicating that 
unattached recirculat- 
ion zones exist above 
the centerl ine down- 
stream of these flames . 



Bonne (1971) 



Two-dimensional hydro- 
carbon-air di ffusion 
flames at normal gravity 



Strong convective effects 
were present, complicat- 
ing the data interpret- 
ation. 



Air inlet velocity 
=6.4 ft/sec; met- 
hane injection vel 
oci ty0.17 -1.2 
ft/sec; fuel-tube 
diameter-O. 19 in.; 
duct diameter-6.0 



DEFICIENCIES 



No quanti tati ve 
information on 
soot , vel oc i ty , 
species, radiation 
etc. 



"EN'S 



Methane and pro- 
pane were used. 
Rectangular 
burners were 
employed . 



The study did not 
reveal quantita- 
tive understanding 
of the phenomena 
of ext i ngui shfnent 
from the point of 
view of the coupl- 
ing between kinet 
ics, soot forma- 
tion, and radia- 
tion. 



Impcrtance 
Of recircul 
a 1 1 on in 
inverted 
flames is 
demonstrated. 
The tip of 
t*e fla^e 
spread out 
and f-e 
flame was 
capped with 
a concave 
upwa rd 

surface, see 
Tig. 2. 



The results 
show that 
radiative 
ex 1 1 ngj ' s h - 
ment occurs 
prior to 
depletion o f 
tie avail- 
able fuel 
ard oxid-:er. 
B a d ■ a 1 1 v e 
trans fen 
was resre- 
s ible for 
cool l ng 
these flames 
to t>-e ro i n 
of visible 
flame dis- 
appearance. 






$V 



QV3&V 



10 



c 
o 



-3 2 

u 

c 



\ 



\ 



F t <^f.r.led Staonaiion Point (-1 i\ / , 

/ 



Noimal-g Flame 



Inverted Flame 




. O-g Flame. 



Da1a 

RH (-1 9) 



2 4 b 

Fuel Flow Rale (ml/sec) 

Fig. 2. Flame height vs. fuel flow rate for a confined gas jet 
diffusion flame. Air inlet velocity = 6.4 ft/sec, 
methane injection velocity = 0.17 - 1.2 ft/sec, fuel 
tube diameter = 0.19 in, and duct diameter = 6 in; 
from SAIC's results. The results show the predicted 
characteristics of normal-g, zero-g and negative-g flames 
along with the experimental data, recirculation, and 
lift-off. 



Prediclions 



11 



presence of buoyancy, the coupling between combustion kinetics and 
soot formation is strong and the present models are not quite capable 
of predicting the effect of buoyancy on the formation of particulates 
Flame radiation is mainly due to the presence of soot particles and the 
interaction between kinetics and radiation is directly influenced by 
their generation. These mechanisms are accentuated in the absence of 
buoyancy. The orange-reddish appearance of the near-zero-g flames is 
an indication of pyrolysis in the hot fuel-rich portions of the flame 
which results in appreciable amounts of solid carbon (Cochran and 
Masica, 1970; Cochran, 1972; Edelman et al. 1973). These studies have 
shown that flame temperature is reduced due to the large globular flame 
surface and continuum radiation from soot. Thermal radiation can re- 
sult in an energy loss of up to 40% of the heat input which augments th* 
kinetic effects that can contribute to flame extinguishment. The 
studies also suggest that pyrolysis and soot formation are of poten- 
tial importance in terms of their effect upon the local molecular weight 
(buoyant force). Experimental results have revealed the non-uniform 
heat-release mechanism throughout the flow field; part i al combustion and 
soot formation dominate in the near-jet region, while the soot burn- 
off process is facilitated downstream due to the increased residence 
time and continued heatir:. The kinetics of the combustion process 
under microgravity conditions significantly influence the accuracy of 
the predictions as can be seen in Fig. 3, which shows the effect of heat 
release on flame shape and its comparison with the experimental data 
The partial oxidation analysis shows smaller flame radii and shorter 
flame lengths, a behavior observed even for the steady-state predictions 
of flames that were actually quenched during the drop-test sequence 
Further results using the partial oxidation analysis models (Edelman 
et al. 1973) are compared with the experimental data of Cochran (1972) 
in Fig. 4; they show substantially improved agreement with respect to 
the results obtained assuming complete combustion model. This tends 
to confirm the potential role of chemical kinetics in microgravity 
flames. 

Ignition, flame stabilization and flame propagation are directly 
related to kinetic effects and are transient in nature. Existing data 



12 




Experiment 

Frozen flow, CH/air only 



.. Full chemistry model 

„H 2 and CO only 
^ H 2 and C(s) only 



Data 



q = 1.08 cm 3 /sec 
Re = 40.9 
T = 294 K 
P = 1 atm 
r = 0.051 cm 



0.4 0.8 1.2 1-6 2.0 2.4 

r(cm) 



2.8 



Fig 3 Effect of chemistry model upon analytical prediction of a 

steady-state zero-gravity methane-air flame; reproduced from 
Edelman et al . (1973). The experimental data were obtained 
from Cochran (1972). 



24 r 

20 

16 



^ 12 



Complete combustion model; 
H 2 0, C0 2 , CO formed^ 



Parti 
theory 



al oxidation 



o 



• * i 



H-0, CO H-0, C(s) 

2 O t) 



r (cm) 



0.051 
0.0825 



Xl L 



Solid symbols denote 
experimental data 



■ ' ' 



10 



100 



1000 



Re fl 



Fig 4 Comparison of theoretical model with experimental zero-gravity 
flame widths in ambient methane-air flames, as a function of 
fuel Reynolds number; reproduced from Edelman et al . (1973). 
The experimental data were obtained from Cochran (1972). 



13 



is insufficient to determine how the combined effects of chemical 
kinetics, soot formation and radiation affect .nese phenomena under 
both normal-g and near-zero-g conditions. As a matter of fact, most 
of the steady, normal -gravity flames of the drop-tower experiments 
were extinguished during the drop (Cochran and Masica, 1970; Cochran, 
1972), suggesting that a combination of chemical kinetics, radiation 
and transient phenomena are responsible for extinction. The major 
role of transient effects is demonstrated in Table 2. There exists a 
certain range of fuel velocity in which a transient period occurs be- 
tween steady state and flame extinguishment. It can be seen that ex- 
tinguished, transient and steady-state flames contribute equally to 
the construction of this table. The parameter characterizing the 
relative importance of transient effects is the Strouhal number 

^ r m /u m T c^ i-e " the ratio of the resi 'dence time to the time associat- 
ed with the drop sequence (see Edelman et al . 1973). Here, r = max- 

r . , . m 

lmum flame radius, u m = axial velocity at maximum flame width and x = 

time required for the flame to attain its minimum length upon the re- 
lease of the test cell. Comparisons between zero-g and normal-g flames 
have shown that the residence times are 5 to 10 times longer under zero- 
g condition. Also, for residence times of the order of 100 msec, sus- 
tained combustion under zero-g conditions has not been observed. The 
methane-air flame data show that when quenching occurs it is generally 
preceded by the attainment of a minimum flame length in times on the 
order of 50-100 msec, which is a characteristic time associated with 
the drop sequence, i.e., the Strouhal number is of the order of unity 
or greater. The hot products of normal-g combustion are retained with- 
in the zero-g flames because of the reduction in convective transport. 
The flame zone is then shielded from fresh oxygen due to the retention 
of the hot combustion gases. This promotes pyrolysis of the constantly 
flowing fresh fuel through the flame. A critical reduction in temper- 
ature due to radiation occurs before the flame is given enough time to 
adjust to a steady-state configuration with convective transport match- 
ing diffusive transport. The result is a reduced heat release rate 
probably due to the change in the oxidation kinetics which ultimately 
results in flame quenching. Figure 5 shows an extinguished flame under 



14 






c 

ro 

i_ 
.E 
O 

o 
o 

E 
o 



-o 
ai 
o 

3 

-o 
o 

1_ 

Q. 

0) 



c 

E 
s_ 

<D 
X 



0) 

O 

I— 
i 

a. 
o 

s- 

Q 

a> 

4-> 

4- 
O 

(O 
■•-> 
(O 

Q 



E 
<0 



0) 

c 



CM 
0) 







15 



3. 1 1 — 



2. 9 



2.7 



2.5 



£ 2.3 



2.1 



19 



17 



4-5 4.7 4.9 



Zero gravity 



5. 1 5. 3 

Time, sec 



5.5 



5.7 



Extinction 



5.9 



Fig. 5. 



Flame length as a function of time for CH 4 /air flames under 
zero gravity Normal -gravity length - 2.78 cm, methane flow 
rate = 2.8 cm-Vsec, nozzle radius = 0.186 cm, velocity = 
25.44 cm/sec, zero-gravity minimum length - 1.86 cm, and 
extinguishment length - 2.93 cm; reproduced from Cochran and 
Masica (1970). 



16 



zero-g conditions (Cochran and Masica, 1970). A flame which (apparently) 
reaches steady state in the 2.2-sec. drop-tower experiment is shown in 
Fig. 6 (Cochran, 1972); whether the oscillations persist given suffic- 
ient time is unknown. The reported steady-state flames may, in fact, 
be in transition from a transient state to extinguishment. However, 
Haggard (1981) has shown that relatively low forced-air velocities 
(= 10 cm/sec) are sufficient to sustain methane flame combustion in 
zero gravity. This experimental result shows the sensitivity of the 
extinguishment process to convective transport in these low-momentum 
diffusion flames. 

We have recently developed an analytical model for the shapes 
and heights of zero-g, laminar, cyl indrically symnetric, transient 
diffusion flames which incorporates the effects of axial diffusion 
(see Sec. IV). The assumptions are: (i) flame-sheet approximation, 
(ii) equal and constant properties, flow velocities and diffusion 
coefficients in the two gas streams, and (iii) negligible buoyancy 
effects, i.e., zero gravity. The only adjustable parameter in this 
analysis is the diffusion coefficient. The results of the mathemati- 
cal representation of flame height as a function of time are in good 
agreement with tnose of the drop-tower experiments for the early por- 
tion of extinguished flames and steady-state zero-gravity flames (see 
Sec. IV). This analysis shows that the theoretical steady-state flame 
lengths are approached slowly. Since the theory does not include the 
effects of radiation and chemical kinetics, quenching can not be pre- 
dicted by this analysis and the predicted approach to steady state may, 
in fact, be faster than would be the case if these effects were taken 
into account. 

The importance of axial diffusion for short flames has been re- 
alized since 1928 (Burke and Schumann). It has been included in the 
analytical solutions for laboratory-scale diffusion flames (Penner, 
Bahadori and Kennedy, 1984) and in the flame height calculations of 
composite solid propellants (Beckstead, 1981) because of the compar- 
able sizes of physical dimensions and flame heights. Observations in 
the drop-tower experiments have shown the existence of much more 
globular-flames compared with those at normal-g with flame height/ 



17 



3.9 
3 8 
3.7 
3 6 
3.5 
3.4 

33|— 
3.2 



Steady slate 

— zero-gravity ■ 
length 



O 

8 
o 

o 

o 



5 3.1 



xz- 3.0 
•& 

■* 2.9 
I 2 81 — 



O 

o 



2.7 
2.6 



_o 



2.5|=S 
2.4 






2.3 
2.2 
2.1 
2.0- 



o e - 



T7 



' OO O oo o 



-Average normal gravity length 



Normal 
gravity - 



<&,' 



,*• Minimum zero- 
gravity length 



± 



-Zero gravity 



III! 



• 6 ~- 4 " 2 ° 2 -4 -6 -8 1.0 1.2 1.4 1.6 1.8 2.0 
Time in zero gravity, sec 



Fig. 6. Flame length as a function of time in zero gravity for Run 3 of 

Table 2; reproduced from Cochran (1972). For this CH 4 /air 

flame, average normal -gravity length = 2.48 cm, nozzle 

radius = 0.051 cm, zero-gravity minimum length = 2.1 cm, 

steady-state zero-gravity length = 3.8 cm, fuel velocity = 

291.2 cm/sec, maximum normal -gravity radius = 0.37 cm, flow 

3 
rate = 2.1 cm /sec, and steady-state zero-gravity maximum 

radius = 0.72 cm. 



18 



maximum flame radius ~1 (see Edelman et al. 1973). In addition, signi- 
ficant deviation between theory and experiment is reported for low- 
Reynolds number flames (i.e., fuel-jet Re~100) in the drop-tower 
experiments. These results suggest that axial diffusion in flames 
under microgravity conditions is as significant as radial diffusion 
especially during the transient development of the flame toward a 
possible steady state and in late-time flame extinguishment. Figure 
7 shows the effect of axial diffusion on both transient and steady- 
state behavior of a methane-air diffusion flame obtained from the 
analytical derivation for unsteady, laminar diffusion flames with 
cylindrical symmetry in the presence of axial diffusion (see Sec. IV). 
This result shows that axial diffusion has a significant effect on 
the development of the flame under zero-g conditions. 

III. Objectives 

The overall objective of the experiment is to gain a better 
fundamental understanding of laminar gas jet diffusion flames in gen- 
eral, and the effect of buoyancy in particular. 

This overall objective will be achieved by meeting a set of speci- 
fic objectives that involve obtaining new data for elimination of the 
inadequacies in the existing data and theory. This will be accomplished 
by studying the following phenomena: 

a) Unsteady phenomena associated with ignition and flame development; 

b^ Steady-state structure; 

c) Soot generation and radiation; 

d) Quenching phenomena. 

Specific collateral objectives involve measurements that include: 

1) Flame shape development and flame extinction; 

2) Color and luminosity; 

3) Temperature distributions; 

4) Species concentrations; 

5) Radiation measurements. 

Thus by eliminating buoyancy-induced convective effects, this 
microgravity experiment will clarify the existing fundamental uncertain- 
ties by: 



19 




! o.o 



R r°- 318 i.o 



WITH AXIAL DIff'JSION 

without a hal oirrrsiow 



D -- 


S.O en 


'/»« 


U = 


10 Cn 


'sec 


V 


315 


cm 


V 


30 cm 




cw < 


/Air 





I 
2.0 



r, cm 



Fig. 7. Comparison between the shapes of laminar diffusion flames in 

both absence and presence of axial diffusion using the analytical 
solution for transient flames (see Sec. IV); D = binary 
diffusion coefficient of CH 4 in N 2 (evaluated at T = 1800 K), 
U = equal and constant velocities of the two streams, R.(R ) = 
inner(outer) tube radius, r = radial coordinate and z = axial 
coordinate. 



20 



i) Obtaining as much data as possible from reduced-gravity 
earthbound experiments; 

ii) Contributing to the development and validation of reliable 
theoretical analyses that form a basis for encapsulating and, 
thereby, interpreting the experimental observations; 
iii) Providing insight into the unsteady phenomena of ignition, 

flame propagation and extinction under microgravity conditions 
with direct applications to fires in space; 

iv) Obtaining required data for future microgravity studies of 
laminar flames, transition from laminar to turbulent flames 
and, consequently, the behavior of fully developed turbulent 
flames by eliminating the complex phenomena cf buoyancy- 
generated turbulence. 

IV. Current Modeling Status 

In this section, we present results of the application of two 
theoretical analyses, each developed to examine certain aspects of the 
mechanisms controlling diffusion flame behavior. The first analysis 
accounts for inertia, diffusion, viscosity, thermal conductivity, and 
arbitrary g-levels for fuel jets issuing into a coaxial air stream of 
arbitrary velocity under steady flow conditions. The second analysis 
accounts for transient and axial diffusion effects under the special- 
ized conditions of constant and uniform velocity and physical proper- 
ties. In addition, an advanced transient, two-dimensional analysis 
and finite-difference methodology which would be adopted for applica- 
tion in the future work is described. 

1. . Steady Flow Jet Diffusion Flame Analysis 

The mathematical model of Edelman et al . (1973) has been used, 
with modifications, to obtain solutions for laminar jet diffusion 
flames. The approach involves a finite-difference solution of the 
partial differential equations for mass, momentum, elements, and energy, 
which are in their boundary- layer form. The effects included are dif- 
fusion, viscosity, inertia, combustion, gravity, and radiation. The 
chemistry assumed is that of shifting equilibrium. 

Figure 8 shows the flow field of an axisymmetric, steady-state, 
vertical laminar jet. The fuel is injected at a finite velocity into 

21 



rMAXIMUM TEMPERATURE SURFACE 
\I"FIAM£ FRONT") 



POTENTIAL CORE RECI0N-, 



JL 



IT •OtMDAMCS OF Tt« 
I V <*FU«MW HEOON 



INITIAL FUEL 
VELOCITY PROFILE 

"I 

INITIAL MASS FRACTION 
DISTRIBUTION 



FOR AIR 



T r T ° T 



INITIAL TEMPERATURE 
DISTRIBUTION 




GRAVITY 
FORCE 



ftAOMTIOfi 



QUIESCENT 
ENVIRONMENT 



0|, WASS 
FRACTIONS) 



1 FOR PURE FUEL 
INJECTION) 



Fig. 8. Schematic of vertical laminar jet; reproduced, with 
modifications, from Edelman et al . (1973). 



22 



s 



the quiescent surroundings. The effects upon the velocity field are 
hear or molecular diffusion of momentum, pressure gradient and grav- 
ity The buoyant force is a combination of gravity and pressure 
gradient. In the vertical jet problem, convection of energy, heat 
release rate, heat conduction, and the energy transported due to dif- 
fusion of species affect the temperature field through the velocity 
fi-ld and the concentration field. The concentration field is, in 
turn, affected by the convection of species through the flow and 
therefore by the velocity field. Chemical production rate and diffu- 
sion are the mechanisms which directly cause changes in the species 
field The diffusional gradients produce an air/fuel ratio distribu- 
tion ranging from fuel rich to fuel lean in the mixing region (see Fig. 
8) The surface upon which the equivalence ratio is unity is referred 
to as the "flame front" which corresponds to the maximum temperature. 
The assumptions made are: (a) diffusive processes are important only 
in the radial direction, (b) combustion is diffusion controlled, and 
(c) flow is steady and axisymmetric. The equations of continuity, mom- 
entum, diffusion of elements, and energy are nondimensionalized with 
respect to the state of the flow at the maximum flame width point (see 
Edelman et al . , 1973). The equations of state are used to complete the 
sot of equations. Mul ticomponent diffusion has been employed in the 
analysis and radiation is currently accounted for in terms of water and 
C0 ? radiation in the simple form: 

where q = radiative heat loss, o * Stefan-Boltzmann constant, e - 
emissivity, X = mole fraction, and T = gas temperature. 

The governing flow field conservation equations are transformed, 
into the von Hises plane, and then solved numerically by employing an 
explicit finite-difference technique. The details of the flow field, 
i e velicity, temperature and species fields along with the gross 
characteristics including combustion lengths and flow deflections have 
been obtained by the solution of the system of conservation equations 
The global continuity equation has been eliminated fro. the system of 
differential equations by introducing the von Mises coordinates as the 



23 



independent variables i p tho +*.= „,-* 

' 1,e " the tran sformat 10 n x, r - x, * is defined 
according to the relations e ° 

-pvr N = ^V x . (2) 

where N - (for plane 2-D flow) or 1 (for axisymmetric flow), p - 

te^and , U I "-'J Vel0CUy ' V = rddl ' al Ve1 ° City ' " - ^ cooniln- 
te, and subscripts r and x denote differentiation with resort t, 

these variables. Figure 9 shows a generic point (n+l.M) in the 
x - * grid network. The finite-difference formulation for the calcu- 
lation of the flow at the point (n+l.M) is obtained by using the 
following explicit difference relations where P i s only one of the 
three pertinent variables axial velocity, element mass fraction or 
mixture stagnation enthalpy: 

i_ P . V 1, M ' p n, M 

dx Ax ' (3) 

If . 1 P n» M * 1 " p n, M - 1 



Af h aP l- b n.M^[Pn,M^-P n .Ml-b n , M -± ft,,^] 
where 



b = pur 2 %/^ N , 
,r VM±i*l[b niM *b niM21 ] f 
f = M(Af) t 



(6) 
(7) 
(8) 



and « is the viscosity. Thus, for example, the momentum eouation at an 
arbitrary point in the field takes on the following W 



24 



M + l 



M — 



M-l 




t>x 



Fig. 9. 



Schematic of the grid network used in the explicit finite- 
difference technique. The arrows indicate the calculation of 
the flow field at point (n+1, H) from data at station n; 
reproduced from Edelman, Fortune and Weilerstein (1972). 



25 



2 

• ♦ b n,M-|]Un, M -" n ,M. ; U„, M .,}.(P e -P) g - ?: A5_. 

(P u )n,M { } 



r 



The remaining conservation equations are described in a similar manne 
in Edelman, Fortune and Weilerstein (1972). However, the inclusion of 
radiation and modified boundary conditions represent modifications of 
the original formulation. Radiation is included via the approximation 
of optically thin medium. The modified boundary conditions provide for 
the presence of confinement to account for pressure field that develops 
in the case of the ducted flow. The effect of flow confinement has 
been included by using (a) skin friction coefficient for shear stress 
calculations at the wall, (b) displacement thickness for effective 
area variations in the viscous layer, and (c) inviscid flow calculations 
for the region bounded by the mixing region and the boundary layer which 
provides the effects of pressure variation due to confinement. 

The present model incorporates the effect of radiation, which is 
an improvement over the previous models. However, other mocifications 
(e.g., soot radiation, detailed kinetics and elliptic effects) still 
need to be included. An interesting result of the model is presented 
in Fig. 10 which shows that in order to eliminate the take-over of 
convective effects under reduced-g conditions, the gravitational ac- 
celeration must be less than 10" 3 g. 

2. Analytical Soluti on for Transient Jet Diffusion Flames 

In this section, we present analytical solutions for the shapes 
of laminar, transient diffusion flames with cylindrical symmetry in the 
presence of axial diffusion. The solution is based on, and is an exten- 
sion of, the previous steady-state analyses (see Burke and Schumann, 1928, 
Penner, Bahadori and Kennedy, 1985, and Bahadori, Li and Penner, 1986). 

The assumptions used in the formulation of the problem are: (i) 
constant pressure, which eliminates the momentum-conservation equation, 
(ii) Lewis number of unity, which decouples the energy- and species- 



26 



= 




Fig 10. Results of the numerical calculations (using the modified model 

of Edelman and co-workers, 1973) for normalized velocity along the 
jet centerline vs. normalized vertical distance for different 
g -levels. These calculations are carried out for a steady-state 
m ethane-air flame. The jet velocity is U Q - 93.5 cm/sec and the 
nozzle radius is 0.0825 cm. The axial distance is normalized 
using a value of L - 50 cm. Note the increase in velocity beyond 
the axial location of z = 4 cm for 10" 2 g (which is a result of the 
the take-over of convective effects). 



27 



conservation equations, and thereby, reduces the problem to one govern- 
ing equation, (iii) equal and constant flow ^.^ transport 

met ers_and physical properties in the two gas streams, (iv) flame-sheet 
approximation, which defines the flame location as a boundary at which 
the fuel and oxidizer have vanishingly small mass fractions and arrive 
m sto 1C hiometric mass flux ratios, (v) no buoyancy effects, and (vi) 
inclusion of both axial and radial diffusion. The assumption of eoual 
(and constant) velocities might seem to be the most limiting one This 
assumption greatly simplifies the analysis because of the elimination 
of the momentum equation. But the effect of velocity on shape is rel- 
atively small as shown in the previous numerical calculations; in fact 
he velocity drops to the ambient velocity in distances (from the center- 
Une) of approximately 2-3 radii of the inner cylinder under normal-g 
condn 10 ns. This implies the domination of diffusion in reoions where 
the flame is present. The assumption of constant diffusion'coefficient 
is more restrictive. However, application of the present model in a 
parametric way has allowed us to extrapolate the transient drop-tower 
data of Cochran and Masica (1970) and Cochran (1972), and obtain esti- 
mates of the time to reach an apparent steady state. The drop-tower 
data have shown that upon the release of the test cell and removal of 
the convective velocity due to the drop in the g-level, the flame 
eight drops to a minimum which is an indication of a tendency to quench 
ut since the fuel is constantly flowing into the products of combustion! 
the flame starts to develop from the new position and (in some cases) 
reaches an apparent steady state. The analysis approximates this event 
by characterizing the turning on of the fuel valve and igniting the 
flame at the same time and recording the flame development. 

For the cylindrical^ symmetric flow of gases, we consider the model 
of Fig. 11 with an inert gas entering along the inner of two concentric 
cylinders (with radius d/2) for times t<0. The oxidizer flows in the 
outer cylindrical shell (bounded by the radii d/2 and d'/2) at all times 
The flow of the inert gas is replaced by the fuel flow of the same vel- 
ocity at times t>0. i.e., the interface between the fuel and the inert 
gas arrives at the port at t = 0, and ignition occurs when this interface 
is at the burner port. This model enables us to mathematically formulate 



28 




inert- 

oxidizer 

boundary 



inert- 
fuel 
boundary 



(a) t<0 




flame 
boundary 
at time 
t 2 (>t ] ) 

flame 
boundary 
at time 

*1 



(b) t>0 



Fig. 11 



Schematic diagram of concentric cylinders with inert I entering 
at z = for 0<r<d/2 during the times t<0 [see (a)] which is 
replaced by fuel F during the times t>0 at the burner port. 
The oxidizer is entering at z = for d/2<r<d'/2 at all 
times [(a) and (b)]. The flame development is shown in (b) for 
two different times. Here, d/2 is the radius of the inner cy- 
linder, d'/2 is the radius of the outer cylindrical shell, z is 
the axial coordinate, r is the radial coordinate, and r f (z,t) 
is the flame radius. 



29 



the problem and obtain an analytical representation of flame-shape 
development as a function of time. 

The species-conservation equations with both radial and axial dif- 
fusion (and no radial or tangential velocity components) for fuel and 
oxidizer above the burner port at times t>0 become, respectively, 



av F 



aY F u aY F n Z 1 d Y F a 2 Y F a J Y F \ x 






(10) 



<r<d'/2,z>0, t>0, ( 



r i 



where Y = mass fraction, u = velocity, D = diffusion coefficient, t = 
time, z = axial coordinate, r = radial coordinate, r f = flame radius, and 
subscripts F and denote the fuel and oxidizer, respectively. Equations 
(10) and (11) reduce to a single differential equation 



av dY /t dY d 2 Y a 2 Y\ , 

which is valid over the entire region 0<r<d'/2 once we impose the 
conditions of u Q = u p = u = constant and D Q = D p = D = constant and 
introduce the single variable 



Y(r,z,t) = 



Y F for 0<r<r fj z>0, t>0, 
i>Y Q forr f <r<d'/2, z>0, t>0, 



(12) 



(13) 



where v is the stoichiometric ratio of grams of fuel consumed per 
gram of oxidizer at a flame front. In view of Eq. (13), the initial 
conditions are 



Y(r,z,0) = 



for0<r<d/2, z>0, t=0 
-"Y , o for d/2<r<d'/2,z>0,'t = 0, 



(14) 



Y(r>0>t) J Y F ; o^0^r< d /2,z=0,t>0, 

'-^O'o for d/2<r<d'/2, 2 = 0,00, 



(15) 



30 



where 
condition is 



the subscript o denotes the initial condition 



s. The symmetry 



dY(0,z,t)/dr = 0, 



(16) 



and the boundary condition is 



dY(d/2, z,t)/dr = 0, ( 17 > 

of oxidizer flow through the wall. The 



which corresponds to the absence 

f lam e-sheet approxi.ation then defi.es the Nation of the flan, bound- 

ary using 

(18) 



Y(r f ,z,t)=0 

which then gives an analytical experssion for flame shape in the for* 
r f (z,t) = 0. 

The method of separation of variables when applied to Eqs. (12)- 
(18) gives the flame location at any time t as 

^, 2 ( Y F,o + *V) (f) °1 (*> f) °» ft ""') 



3 = 2 



M *2 



L(f 



x exp 



(W) 



2Dt 



4Dt DA<t + in 




1 - 1 + 



/ Y F.o + vY 0,o\ 
\ vY 0,o 11 



z\2 



J (t * !) 



■ -[('f.. + ^ Y o,o) (!■) 2 - * ¥ °.o] 



x exp 



«(»♦!) 



2Dt 



'1 



Y r . (d/d') 2 



4Dt *n 
1 + 






(d/d 1 ) 2 ] 



- 2 (t ♦ !) 



z\2 



z\ > 



(19) 



J 



31 



where J Q and J ] are Bessel functions of the first kind of orders 
zero and one, respectively, and the eigenvalues A. are the roots of 
^(X^'/2)- with \. f 0. For large times (i.e."! t - ») , E q. (19) 
reduces to the following solution for steady-state flame shape: 

h Mff^j exp I* ( ] -r + ~j~)[ 



-[( Y F,o * vV 0>o j (^ - , Vo J 



In the absence of axial diffusion, Eq. (20) reduces to the classical 
solution of Burke and Schumann (1928) for steady-state, cylindrical^ 
symmetric laminar diffusion flames with no buoyancy (see Penner, 
Bahadori and Kennedy, 1986), viz. 



Equation (19) was solved numerically by using polynomial expres- 
sions for Bessel functions. The input to the problem consists of the 
values for Y^, v , Y^, d/2, d'/2, u, D, and a selected number of 
times ranging from 0.05 sec. to 100 sec. The radial distance d'/2 was 
divided into 40 intervals which were not necessarily equal due to the 
steep variations in mass fractions. The number of eigenvalues calcu- 
lated in the program was 2000; these were obtained by finding the roots 
of Jjttjd'^) = o. A large number of eigenvalues were needed (espec- 
ially at smaller times) due to the slow decay of the exponential terms. 
Calculation of the terms in the summations were stopped once the expon- 
ential terms became smaller than 10" 6 since the coefficients in front of 
the exponentials are always of the order of 1 or smaller. Once all the X 
were calculated, J (A.d 7 2) and J^d/2) were obtained. Provisions were" 
made to calculate J o (y) at the beginning of the program for any sel- 
ected radial distance r. The steady-state flame height calculations were 
performed by setting r = in Eq. (20) and obtaining a value of z for 



.i 



(20) 



9 /<L\y> J l( A J I) J o ( A i ^fjj xP' (-P*j z /") vY n IA 2 

Id'/Zrf / d'\ 2 / — dV — ~ -=v Q.o . i_\ . ,o! 

J-2 ( A j M J o 2 (*jf-) Y F.o + " Y 0.o Id') (21 



32 



which this equation was satisfied (using marching in the axial Erec- 
tion from the burner port for specified step sizes.) The steady-state 
flame height served as an upper limit for the axial locations along 
the flame boundary in the calculation of steady-state flame shape, 
transient flame heights and transient flame shapes. For each select- 
ed axial location, one point in the radial direction was obtained at 
which a sign change in mass fraction occured; this is an indication of 
going from fuel region to oxidizer region, i.e., location of the flame. 
The time-dependent flame heights and shapes were then obtained by fol- 
lowing the procedure for calculating the steady-state flame shapes but 
using Eq. (19) for any selected time. 

The results of the present model are compared with the data ob- 
tained from the droo-tower tests in Figs. 12 and 13. Although the as- 
sumptions used in the development of the model are somewhat extensive, 
the results are a useful tool for extrapolating the drop-tower data 
and helping us to estimate the time required to reach steady state 
under reduced-g conditions. 

:L_T ransient, Na v ier-Stokes Based Modej 

The model which will be used in the future work consists of the 
modification of an existing SAIC transient, axisymmetric Navier-Stokes 
based model and code that will predict flame ignition and development 
to steady state, include both axial and radial diffusion, include soot 
formation and disposition chemistry, account for radiation from soot, 
CO, and H 9 0, predict flame extinction, and obtain temperature, veloc- 
ity and species fields. The modified model will be in transient, axisym- 
metric, full Navier-Stokes form, which includes two-dimensional conser- 
vation equations for mass, momentum, energy, and species, with arbitrary 
g-levels and both Fickian and multicomponent diffusion. A radiation 
model such as the discrete ordinates model will properly take into 
account the effect of soot, C0 2 and H 2 0, and sub-global kinetics with 
finite reaction rates will be included. Postulated detailed reactions 
involving many active species are available. However, to try to under- 
stand the essential features of diffusion flames, simpler chemistry is 
more appropriate, e.g., two-, three- or four-step global models which 



33 



4.0 " 



•3.0 



?.0 



0.0 




1970) 

Transient, elliptic analytical model 
' - Theoretical steady state 



0.5 



1.0 



Time, 



1.5 
sec 



?.0 



!.S 



Fig. 12. 



Comparison between the analytical and experimental flame height- 
time variations under micro-gravity conditions. Here, D = binary 
diffusion coefficient of methane in nitrogen, U = fuel velocity 
R = nozzle radius, and h^ = steady-state flame height. The value 
of D = 1 cm /sec corresponds to a temperature of 750 K. The 
experimental results show the flame extinguishment at approximately 
0.6 sec (see Fig. 5). 



. 3.0 



£2.0 



0.0 






- 0.80 cir. '/sec-y 


\s ' 4 -"« 


T-Z 


... 0.85 -a 


3.94 




1 .00-y 


3?5 



R = 0051 cm. u « 291.2 cm/sec 

t/psrimc-nt (Cochran, 1972) 

Transient, elliptic analytical model 

-- Theoretical steady state 

'■•■■■ ■■ ' ■ - 



0.5 



10 1.5 

Time, sec 



2.0 



2.5 



Fig. 13. 



Comparison between the analytical and experimental flame height- 
time variations under micro-gravity conditions. The symbols are 
defined in Fig. 12. The value of - 0.85 cm 2 /sec corresponds to 
a temperature of 650 K, The experimental result was reported as 
having reached steady state in 0.4 sec (see Fig. 6). However, note 
the oscillatory behavior in the data and the theory that shows 
steady state has not been reached. 



34 



are capable of delineating the major regions of the reaction process, 
and also lend themselves to the prediction of ignition and stabiliza- 
tion characteristics in detailed numerical computations. 

The numerical technique involves a control volume differencing 
scheme, i.e., the governing equations are expressed in integral font, 
for a finite-difference grid element. A generalized coordinate system 
is used to resolve irregular and moving boundaries. This consists of 
general quadrilateral cells which allows grid points to move with the 
fluid (Lagrangian), be held fixed (Eulerian) or be moved in any pre- 
scribed manner. An acoustically implicit solution for pressure is 
employed to relax the Courant condition on the time step (i.e., a 
semi-implicit method); this is especially suitable for low-speed flows 
which are indeed of interest in the present problem. The staggered 
grid system consists of cell-centered scalar quantities and node- 
centered flow velocities. The stress terms in the governing equations 
are obtained through explicit time differencing and centered spatia 
differencing. For all but the pure Lagrangian solution, the spatial 
advective derivatives are written such that a purely centered to full 
donor cell differencing may be employed. This allows the use of numer- 
ical stabilization without the addition of an explicit artificial vis- 
cosity. 

Within the framework of this model, phenomena including ignition, 
flame stabilization, flame development as a fuction of gravitational 
acceleration, radiation, and environmental conditions can be analyzed. 
Specific attention will be given to the modeling of finite-rate oxida- 
tion, soot formation and combustion processes utilizing global reactions. 
The results of this analysis will provide fundamental information on 
conditions relevant to flame development and flame extinction phenomena 
for which no comprehensive theoretical analysis of the type described 
here currently exists. 

V. Data Necessary to VaU HatW Improve Model (si 



The reduced gravity data that is required to validate and improve 
the existing theoretical models consists of temperature distribution, 
species concentrations, velocity distribution, flame development rom 
ignition, flame spread, luminosity, color, soot formation, and radiation, 



35 



In Sec II the current understanding „ f diffus1orl f , 
shown to be incomplete. The strono convective eff^K -7 
^ -er norma,, -n^m^ 

rr;™r redkt the -— - —'» «- 

The NASA-Lewis drop-tower observations show that sio„if icant 

n 9 es ,n name structure occur durin g the drop. ,„ ma y 
e ta ass0cjated wUh ue tpans(tiM f ^ ^ _ 

m o e ; t;:;: -9 ri tion pers,sts th — «* •»»•"• *-« 

TM, may b to he ,* "T" "! "'" qUenCh " dUHn9 ^ '"">■ 

Such an occurence could lead to quenching or possibly a sL ! 
ment to an "apparent" steady-state zero a fl»l J " St " 

-o P -tower observations. ^J^l ' Zl 7Tt!^ 
mate the uncertainty as to whether in f, ♦ m " 

Nation on ignite under mjcr0gravity ~ ^ ^ ^"^ "" 

- ^ir ::r of tanperature - sp « ies — »- 

- -*,. W ii, e;a:;: p „v;: ::;:;;: :;:;: pa ;r ; h ? numer - 

-tlon. radiation and soot formation. As w " d , " ' nf ° r " 

£ "jrr r than d — «-- -« - „~ 

8 U10n after the release of the test cell in +h« o o 

n,on flame behavior that win be used to finalize the test mat x 
d measu re.ent program for the low-gravity experiments, as scT 

1 beZ " C h * antiUtiVe d " a (e ' 9 - «™«™ »- 1) 

s t " ^^ ° f th6Se " Pen ' me " tS - " *« «- '^ta- 

rn n a T" a Pr ° D,em ,n ™ etin9 "" CrUical "'•«» retire- 
ments, a P!an for a space experiment win be developed to obtai these 



36 



data. 

In general, the data obtained from the experiments will be used 
in the development and validation of the theoretical analyses. Table 
3 and Fig. 14 show the relationship between the various data obtained 
from the experiments and the theoretical modeling. 

VI. Identification of Experiments 

A total program of experiments for laminar gas jet diffusion 
flames is given in Tables 4 and 5. The experimental techniques that 
may be applied are cinematography, sampling, temperature and pres- 
sure measurements, accelerometry (if accelerometers are available), 
and radiometry. The parameters that can be varied during the entire 
program are: chamber pressure, type of fuel, nozzle diameter, fuel- 
flow rate, and chamber oxidizer (i.e., 2 /N 2 ratio). The fundamental 
effects that can be studied are: transient behavior, igniton, chemis- 
try, soot yield, radiation, fuel reactivity, and diffusi vities. 

A matrix of tests has been selected based on the reouirements 
summarized in Table 4. This matrix consists of 48 flames which in- 
clude three flow rates, two nozzle radii, two fuels, two chamber ox- 
idizers, and two chamber pressures. Three flow rates are selected 
because the flow rate is an important parameter governing the struc- 
ture of the gas jet diffusion flames. Two nozzle diameters are selec- 
ted to provide independent variation in Reynolds number for a given 
flow rate where the Reynolds number is shown to be important with 
respect to the degree of the transient behavior observed in the flame. 
The combination of two pressure levels and two environmental composi- 
tions is sufficient to define the effect of oxygen partial pressure 
on flame chemistry. Methane is selected to be the primary fuel be- 
cause of the available background data and theory. Velocities will be 
determined from the theoretical analyses while radiation measurements 
will orovide an indirect measure of soot formation. 

Upon completion of preliminary testing including ignitor develop- 
ment in the 2.2-sec. drop tower, the complete matrix outlined above 
will be carried out utilizing both the 5 -sec. drop tower and the 
specially outfitted KC-135. 



37 



Table 3. Relationship Between Data and Modelinc 



COUPLED EFFECTS: INERTIA, CONVECTION, DIFFUSION, VISCOSITY, 
CONDUCTIVITY, RADIATION, CHEMISTRY, GRAVITY 



FUNDAMENTAL 

UNCERTAINTIES 



EXTINCTION 



IGNITION 



FLAME 
DEVELOPMENT 



OSCILLATORY 
BEHAVIOR 



SOOT FORMATION 
AND DISPOSITION" 



KEY 
MECHANISMS 



KINETICS, RADIATION 



KINETICS, CONDUCTION 



KINETICS, DIFFUSION, 
CONVECTION 



TRANSIENT, INERTIA 



KINETICS 



DATA" 



MODELING* 



SPECIES, TEMPERATURE, 
VISUAL, RADIATION 



VISUAL 



TEMPERATURE, SPECIES 
VISUAL, RADIATION 



VISUAL 



TEMPERATURE, SPECIES, 
SOOT, VISUAL, RADIATION 



TRANSIENT, 2-D 
NAVIER-STOKES 
INCLUDING CHEM- 
ICAL KINETICS 
AND RADIATION 



•CONNECTION BETWEEN DATA AND MODELING: PREDICT TEMPERATURE, SPECIES, VELOCITY FLAKE-SHAPE 

!SSKlSs c : and compare with measured results (EXCLU ° ing v ^ [ ™ t ° ™ 2S 

I USE SIMPLISTIC MODELING, ADD RFSULTS TO THF nATA BASE. 



38 




FXPFR1MFNT 



TEST MATRIX 



5-SEC. DROP 
TOWER AND 
KC-135 
RESULTS 



THF(lRFTir fll amaivsis 
MO DELING 1 



PARAMETRICS 



COMPARISON WITH 



PRED1C 



SUGGEST SPECIFIC 
TESTS 



IONS 



MODEL REFINEMENT 



ANALYSIS AND 
DATA BASE 



D 



Fiq. 14. Relationship between data and modeling. 



39 



Table 4. Operating Conditions, Ranges of Variables and Applications to 

Model Development for Defining the Total Program of Experiments. 



OPERATING 
CONDITIONS 



Fuel Flow Rate 



Nozzle Radius 



Fuel 



Chamber 
Pressure 



Chamber 
Oxidizer 



RANGES OR 
TTPES 



"Ch 



I .0, 2.0 and 
4 3.0 cc/sec 



Or h = °- 5 . '-0 and 
3*8 1.5 cc/sec 



0.051 and 
0.0825 cm 



APPLICATIONS TO MODEL 
DEVELOPMENT 



DISCUSSION 



Effect of injection velocity 
on flame characteristics 
(for a given nozzle radius) 
to help sort out the effects 
of diffusion processes. 



Methane; Propane 



P. » S a tm 
and 1 atm 



(02/N 2 ).0.21/ 
0.79 and an 
oxygen-defi- 
cient O2/N2 
mixture (com- 
pared to air) 



Effect of no2zle size on 
flame characteristics (for 
a given fuel-flow rate) to 
help sort out the effects of 
diffusion processes. 



Effect of Reynolds number 
on steady state, transient 
and quenching phenomena can 
be studied. 



Effect of fuel reactivity, 
stoichiometry and diffusivity 
on flame structure and be- 
havior, kinetics, and soot 
formation. 



Effects of pressure on soot 
formation, quenching, rad- 
iation, kinetics, and Ign- 
ition. 



CHa Is selected because of 
the base of data that al- 
ready exists. C3H8 is sel- 
ected to examine a heavier- 
than-air fuel, and because 
it is representative of a 
wide range of higher-mole- 
cular weight hydrocarbons. 



Effects on the kinetics, soot 
formation and ignition be- 
havior. 



Reduced pressure is used to 
enhance kinetic effects. It 
indirectly clarifies the 
effects of lower O2/N2 ratio 
(compared to air) via partial 
pressure. 



To study ambient composit- 
ional effects on flame 
structure. A1r (02/N 2 -0.21/ 
0.79) Is selected as a base- 
line. Reduced levels of O2 
are selected to enhance 
kinetics. 



NUMBER OF 
EXPERIMENTS 
FOR TOTAL 
PROGRAM 



£' 48 



40 






Table 5. Measurement Techniques, their Rationale and 
Applications to Model Development. 



MEASUREMENTS 



Cinematography 
(using High-speed 
movie cameras) 



Thermocouples 
(using rakes) 



Sampling probes 

(using sampling 

bottles) 



Pressure 
Transducer 



Velocity 
Measurement 



Radiometery 



Accel erometer 
(if available) 



RATIONALE/APPLICATIONS TO MODEL DEVELOPMENT 



Obtaining time-resolved visualization data on 
ignition, flame development, color, luminosity, 
and quenching. Flame shapes can be compared 
directly with model predictions. 



Obtaining temperature distributions in the flow 
field aoove the flame. These temperature 
distributions can be compared directly with the model 
predictions of thermal field. 



Obtaining species distributions above the flame 
which can be compared with the model predictions 
of species field. 



Measurement of the Increase in chamber pressure 
due to the burning process yields estimates of the 
overall heat release rate. 



Measurements for the scale of this experiment require 
a laser-based diagnostics, which is not feasible at 
this time. This requires a compromise in which 
predictions and other measurements will be used to 
provide the velocity. 



Isolation of thermal radiation, and as a result, 
determination of the effect of radiation on reduction 
in temperature within the flame. Potential exists 
for determination of soot concentration and genera- 
tion Evidence shows that radiation from soot is not 
dominated by C0 2 and H ? 0. Therefore, calculations for 
CO? and H?0 combined with measured radiation yield the 
contribution from soot, and hence, soot concentration. 



Measurements of the gravitational acceleration 
enables us to interpret the data that could be , 
affected by gravitational levels greater than 10 g. 



&&%& 



41 



The present program focuses on laminar flames. Turbulent dif- 
fusion flames will be deferred for the future. 

VII. Apparatus and Conditions 

Experimental conditions, components of apparatus, ranges, require- 
ments, and specifications are presented in Table 6. A generic chamber 
design is shown in Fig. 15. 

Ground-based testing of the experimental chamber, camera, ignition 
system, thermocouples, pressure transducers, fuel tank, linings, valves 
radiometer, sampling probes, and sampling bottles should be conducted in 
order to ensure that: 

a) The equipment operates reliably; 

b) Reproducible results can be obtained; 

c) Calibration of the flow system is accurate 

d) The ignitor functions properly, that flames can indeed be 
ignited and ignition duration is defined (using the drop- 
tower facility) ; 

e) Tubes, valves, fuel tank, camera, thermocouples, sampling 
probes, sampling bottles, radiometer, and the experimental 
chamber can withstand the impact in the zero-gravity facil- 
ity and KC-135 flight conditions} 

f) There is no leakage from the experimental chamber, tubes, 
valves, and fuel ; 

g) Stress analysis on the chamber, tubes, and fuel tank have 
been done, especially for the chamber to withstand the 
highest temperatures and pressures reached. 

VIII. Data Anal ysis 

The films will be developed and analyzed using a motion picture 
analyzer to observe anti measure the flame development and other pro- 
cesses. The thermocouples ' output is analyzed along with the records 
of the chamber pressure. Acceleration records (if available) during 
the experiment will be analyzed, and observations related to micro- 
gravity pertrubations will be explained in the analysis and interpre- 
tation of data and flame behavior. The gas environment in the chamber 



42 



Table 6, Apparatus and Conditions. 



COMPONENT 



Fuel 



Experimental 
Chamber 



Chamber 

Environment 



Nozzles 



Ignition 



Fuel Tank 



REQUIREMENTS 



0-1-3 cm 3 /sec for methane; - 0.5 - 
1.5 cm 3 /sec for propane; pure, high quality 
fules. 



Closed, sealed, either cylindrical or rectang- 
ular. Minimum height/diameter » 1.5. Must 
withstand pressures of up to 3 atm, gas temp- 
eratures of up to 400K and metal temperatures 
up to 310K. 



Two compositions [2U 2 - 79% N ? (air) and 
oxygen-deficient 2 /N2 mixtures Tcompared to 
airj] and two pressures [1 atm (14.7 psia) and 
0.5 atm] for different flames Before each 
test, the gas temperature should be 295 i w. 
Pressure tolerance U0.1 psia); 2 tolerance 
U0.1 mole t at the beginning of the test setj. 
Specifics will be determined from the final 
design of the experiment Including the Input 
from 5-sec drop-tower studies. 



Radius • 0.051 and 0.0825 cm; distance from 
the tip to the chamber ceiling 230 cm; 
di stance Between nozzles • 5 - 10 cm (if more 
than one nozzle per chamber), equally distant 
from the centerline of the chamber and 1n 
line with the camera(s) with the smaller-size 
nozzle closer to the camera; minimum distance 
from wall for nozzle positioning • 10 cm. 



Each nozzle (if more than one nozzle 1s used 
1n the chamber) needs one ignitor Spark 
ignition 1s preferred. The details of the 
Ignition system will be worked out after tht 
test results of the 2.2-sec drop tower become 
available. 



Internal pressure: 
KC-135. 



TB0 for drop tower and 



SPECIFICATIONS AND DISCUSSION 



The amount of fuel required for a 
single test 1s based on the fuel flow 
rate and an estimated time of the 
experiment for each flame. 



The chamber must be able to withstand 
the specified conditions for the 
required period of time. Provisions 
for thermocouples. Ignitors, trans " 
ducers and viewing ports must be made 
1n the chamber. Fig. 15 shows a 
generic chamber design. 



Flames are going to be selected in such 
a way that the maximum amount of oxygen 
burned at the end of the flight does 
not exceed 10% of the original number uf 
moles of oxygen. Chamber pressure should 
be recorded before, during and after 
the experiment. Chamber temperature 
must be monitored over the entire experi- 
ment. 



This nozzle arrangement ensures that 
temperatures <3S0K and negligible gas 
velocities exist at the top of the 
chamber, which prevents both extensive 
heating of the chamber ceiling and 
significant recirculation which would 
otherwise be a source of flame disturb- 
ance. Since the maximum flame radii 1s 
-0.5 - 1.0 cm, proximity of the nozzles 
wlil not cause flame disturbance. 
Nozzles must be secured inside the 
chamber to avoid misalignment or loosen- 
ing due to vibrations 1n KC-135 or im- 
pact in drop tower. 



Ignition is the most crucial factor In 
the experiment. Tests should be con- 
ducted 1n the 2.2-sec drop tower for 
all of tht flames to be studied to 
ensure that reliable and reproducible 
Ignition occurs for each flame. The 
procedure is to drop the test cell, and 
then simultaneously open the fuel valve 
and Ignite tht flow under the near-zero- 
g condition. 



The tank must contain enough fuel 1n 
order to avoid large pressure drops at 
the end of a series of tests per flight 
1n the KC-135. 



43 






Table 6 iCo.it. ). 



COMPONENT 



Viewing 



Fuel-Tubing 
System 



Chamber Cooling 



Acceleration 



Camera 



REQUIREMENTS 



The field of view must be 5-cm horizontal by 8-cm 
vertical at the plane passing through the flame 
centerline and perpendicular to the axis of the 
camera. Besides viewing port for camera, a 
viewing section must be provided for the radi- 
ometer and the experimenter. 



The tubes can be made from stainless steel. The 
accuracy of flow control should be within ♦ 0.01 
cm J /sec. Flow valves, explosion-proof solenoid 
valves, and regulators should be provided and 
tested. Pre-calibrated precision valves might be 
required to maintain specific flow rates. 



SPECIFICATIONS AND DISCUSSION 



Two viewing ports might be required. 



Flow recording during the tests will be 
helpful in ensuring that the required 
flow rates are maintained. 



Times of up to 5-10 minutes are required for the 
chamber environment to cool down to the ambient 
temperature after each flame measurement. The 
maximum gas-temperature rise inside the chamber 
will be less than 80 C (assuming no heat transfer 
to the container) during each test. This maxi- 
mum gas-temperature rise will raise the tempera- 
ture of the container by <2 C if all of the 
heat release is transferred to the metal. 



The acceleration should be recorded at intervals 
during the KC-135 flights If the capability 
exists. The acceptable g-levels are in the 
range 10 3 g - 10" B g. The results of the g- 
jltter analysis show that frequencies >1 HI 
are acceptable. 



Cooling can be facilitated by providing 
an Internal fan for forced convection 
inside the chamber. 



Since the flame development at the early stages 
of ignition is quite rapid, the movie camera 
should be capable of operating at a framing 
rate of approximately 400 frames per second. 
Color, l&ffli film will be adequate for the 
experiment. A timing system should be 
provided to record accurate time Intervals (of 
the order of 0.1 sec) on the film. 



The acceleration should not be greater 
than 10" 3g, because (a) the velocity of 
the burned gas starts to increase down- 
stream after reaching a minimum (see 
Fig. 10) for gravitational accelerations 
greater than 10" Jg, and lb) the gas 
temperature decays more slowly for 
gravitational accelerations greater than 
10* 3 g (see Fig. 16). Burning will be 
affected by g-level (Figs. 10 and 16) 
but measurements will not be adversely 
affected by various g-levels. 



Filming begins prior to the opening 
of the fuel valve and the ignition pro- 
cess which are synchronized. 



ORIGINAL PAGE « 
OF POOR QUALITY 



44 



Table 6 (Cont.). 




Sampl ing 



Quenching quartz microprobes m 



Scaling 



Thermocouples 



Recording 



Radiometer 



^; C "'he r *n«rr.d.t S-10 cm above the burner 
o"r!; and it . ^«UUd ti.e ur-ng h tst.l 1 

volume and final pressure of ~ 5-7 psia. 



-:;; J 'r;;.^:;"r,::;".:;:. t r::r,;* 



A rake of thermocouples -ill be used tomeasure 
temperature distributions In.t e o- el m 
'rr !^ "neS ° m 1 in POlntK with -ire dia- 

be exposed . 



Th» data collection -ill be comprised of film 
! olor 16 m] history of the combustion pro e s 
^e\he discussion «-«« • "jnjocou .( J^ 



o 
pres 



(if the capability exists |. 



A radiometer -ill be designed to record flame 
radiation. 



Perturbations due to sample removal 
donot cause major flame disturbance 
,, long as suction 1s controlled over 
a short period of time and 1s not 
sudden. 



Gas-temperature measurements -il be con 
ducted at fixed locations above the flame 
at 5-15 cm above the nozjle(s). Since 
corrections for heat conduction radiation, 
etc. are available for large P r ° b «- 
accuracy will not pose a major problem. 
The rake -Uh 6-9 thermocouples -ill pro- 
vide temperature distributions 



Other specifics are described in the in- 
dividual sections in this table. 



Specifics -ill be determined from final 
delion of the experiment, complexity of 
the experimental package and test require- 



the expe 
ments 






& 



o* 



? i 



45 




NOTE: Mounting 
Hardware 
Not Shown 



Cylindrical 
Chamber 



Viewing 
Ports 



Fig. 15. Generic experimental chamber design for gas jet diffusion-flame 

experiment. The chamber can be either cylindrical or rectangular. 



46 



£ •••■ 




Fig. 16. 



Results of numerical calculations for normalized temperatures along 
the jet centerline vs. normalized vertical distance for different 
g-levels. Here, T_ x is the maximum temperature reached at the 
centerline and T Q - 300 K is the initial gas temperature. Other 
symbols and conditions are defined in Fig. 10. 



47 



will be analyzed to obtain information in relation to hLO, , CO, 
CO., H 2 , formaldehyde, and hydrocarbon concentrations. 

Theoretical predictions will be used to support the analysis and 
interpretation of the data base developed from these experiments. The 
results of temperature and species measurements combined with the 
global characteristics of the flame (e.g., shape and luminosity) pro- 
vide both quantitative and qualitative information for improving the 
theoretical models. The information on thermal and chemical fields 
will help to gain insight into the chemistry and fluid dynamics of the 
process. When experimental data is used in conjunction with the pre- 
dictions, model validation is facilitated in relation to ignition, 
flame propagation, kinetics, and radiation effects. Also flame-scaling 
criteria will be put on a firmer fundamental basis. 



48 



IX. References 

Altenkirch, R.A., Eichhorn, R., Hsu, N.N., Brancic, A.B. and Cer.Ho,. 
N E (1976) Characteristics of laminar gas jet diffusion flames under 
the' influence of elevated gravity. Six teenth Symposium (Internanona 
2!lI ™busti=n, PP. H65-1174. The C^bustion Institute. Pittsburgh, PA. 

Astavin, V.S. and Ryazantsev, Yu. S. (1979) Temperature and concentra- 
tion distribution in the reaction region of parallel flows of unmixed 
reactants. Fluidjffiamics 14, 274-279; Translated from U^aMademn 
Ito.it SSSR - He """' 7hidkosti i Gaza 14, 139-146. 

Bahadori, M.Y.. U. C-P., and Penner, S.S. (1986) Two adjacent, coupled 
Uminar diffusion flames with cylindrical sy^etry. P^res^njstro^ 
^^^^jeronautics^V.!! HIT Dyn^r. of Reactive Systems, ar _jl 
^^T^TSStions (J.«. Bowen. J.-C. Leyer and R.I. Solouk* , 
Eds.). PP. 192^07. American Institute of Aeronautics and Astronautics, 
New York. 
Beckstead, M.W. (1981) A .ode! for solid propellant combustion ilojjteenth 

^^^^ ,L., Beckstead H.W.. 

N.S. (1969) Combustion tailoring criteria for solid propellants. AFRPL- 

TR-69-190 (835-F), Redlands, CA. 

Bonne, U. (1971) Radiative extinguishment of diffusion flames at zero 
gravity. Combust. Flame 16, 147-159. 

Burke, S.P. and Schumann. T.E.M. (1928) Diffusion flames. Ind^Eno^ 
Chem. 20, 998-1004. 

Cochran, T.H. (1972) Experimental investigation of laminar gas jet dif- 
fusion flames in zero gravity. NASA TN 0-6523, 26 pp. 

Cochran, T.H. and Masica, M.J. (1970) Effects of gravity on laminar oas 
jet diffusion flames. NASA TN 0-5872. 



49 



Edelman R.S. and Bahadorj> „.„_ (]986) EffKts Qf ^ 

J.t diffusion Ha.es: Experiment ,„d theory. Acta Astronaut,, (,„ 



press) . 



Edelman R.B., fortune, 0., and Weilerstein, G. (1,73) Analytical study 
of gravity effects on laminar diffusion flames. NASA CR-,20921, 136 pp. 

TZTn^TT' °; F - wei,erstein - G - c ° chran - t - h - «"* -«-• 

fee s *"'"' a " d ""-'■'-'■f 1 instigation of 

cts pon ammar g as jet diffusion fia.es. F^ne^^Jum 

(Internationa l) on Combustion pp . 399-412 Th» r™,h . • , 

^— —— ^ PP. 412. The Combustion Institute, 

FW, W. L . and Bo^an, C.T. (,983) Measurements of the structure of 

8776. Send,. National Laboratories, Livermore, CA 94550; paper Western 
States Section of the Combustion Institute, WSS/CI 83-56. 

Haggard, j.„. ( 1981) Forced and natura , ^^ ^ ^.^ 
diffusion flames. NASA Technical Paper No. 1841. 2, pp . 

Haggard O.B. and Cochran, T.H. (1973) Hydrogen and hydrocarbon dif- 
fusion flames in a weightless environment. NASA TN D-7165, 28 pp. 

Kimura, I. and Ukawa, H. (1961) A ^t„rf u n r n.. .. -. 

• "■ n«l| A study of the combustion of laminar 

, J ' j apa " S ° c1e ty _oOtecJijnj«lJn a l J neers 27, 736-746- trans- 
lated as NASA Technical Trans.ation, NASA TT F-,3, 459, February ,971. 

Klajn, „. and Oppenheim, A.K. (1982) Influence of exothennicity on the 
shape of a diffusion flame. N ineteenth S^posium f mt^.^^ 
Combustion, pp. 223.235. The Combustion Institute. Pittsburgh. PA. 

Mitchell, RE., sarofim. A.F.. and Clomburg, L.A. (1980) Experimental 
an numenca investigation of confined laminar diffusion flames. c- 

bust. Flame 37, 227-244. - JSL - 



50 



Permer S S., Bahadori , M.V., and Kennedy, E.M. (1984) Laminar dif- 
fusion' flames with cylindrical symmetry, arbitrary values of diffusion 
coefficients and inlet velocities, and chemical reactions in the ap- 
proach stream. yaaregJILirtmn.i.ttr, and Aeronautics, Vol^ 
n^ic^ofJ^exindJ^Oi^Ji^s (J.R. Bowen, N. Manson A K. 
^— -^Zr^^r^l PP. 261-292. Amencan Inst,- 
tute of Aeronautics and Astronautics, New York. 



61 



3fi£tf0JL3 

-Effects of Buoyancy on Gas-Jet Diffusion Flames' 
- Conceptual Design Review- 



R. B. Edelman and M. Y. Bahadorl 



June 1986 



Science Application* International Corporation - 



CONCEPTUAL DESIGN REVIEW 

EFFECTS OF BUOYANCY ON GAS JET DIFFUSION FLAMES 

Raymond B. Edelman (Principal Investigator) 
M. Yousef Bahadori (Assistant Investigator) 

SCIENCE APPLICATIONS NTERNATIONAL CORPORATION 
Chatsworth, Calfornta 

Sponsor 

NASA - Lewis Research Center 
Cleveland, Ohio 

Contract NAS3-22822 

Sandra L. Olson (Project Monitor) 

JUNE 1980 



Science Applications Hamational Cav***" - 



BACKGROUND 



I — Sdanc* Applications ktomationilCapontian- 



WHY THE laminar CAS JET DtFFUS/ON FLAME? 




• EMBODIES MECHANISMS FOUND IN NATURAL 
FIRES AND CONTROLLED COMBUSTION 
PROCESSES 

• SIMPLE AND BASIC 

• EASY TO CONTROL 

• LAMINAR FLAME NEEDS BETTER UNDERSTAND- 
ING AND IS A STEP TOWARD UNDERSTANDING 
THE MORE COMPLEX TURBULENT DIFFUSION 
FLAME 



FLAME STABILIZATION 



I — SriwCTAflptoiw^w i ijrirfCMiwwtai- 



APPROACH 



• ISOLATE THE EFFECTS OF BUOYANCY BY 
ELIMINATING GRAVITY 

• EXPERIMENTAL 

• THEORETICAL 



RESULTS APPLICABLE TO FIRES ON EARTH AND IN SPACE 



Sdmo»AppkeMtiomir«anwUon»l Cor po rat i on- 



OBSERVATIONS FROM PROP-TOWER EXPERIMENTS 



UNDER REDUCED-GRAVITY CONDITIONS FLAMES EXHIBIT 
DISTINCTIVE CHARACTERISTICS RELATIVE TO 
NORMAL-6 CONDITIONS 

. . BECOME GLOBULAR 
. . MORE SOOTY 
. . TEND TO QUENCH 
DROP-TOWER RESULTS HAVE SHOWN THAT 

. . SOME FLAMES EXTINGUISH 
• • SOME ARE IN TRANSITION BETWEEN 
EXTINGUISHMENT AND STEADY STATE 
(LACK OF ENOUGH TIME) 
. . OTHERS REACH (AN APPARENT) STEADY 
STATE WITH FLAME-HEIGHT FLUCTUATIONS 



. Scienee Applications k fo iiW i onalCcipoilt i an- 

r Z2-SEC. DROP-TOWER EXPERIMENTS 




m».» Sm* 



• LOtur*. 



Wm im w y *y . 



Burner radus • 0.051 cm, 
Average normal-g length x 2.48 cm, 
Maximum ncrmaHg radus = 37 cm. 
Methane flow rate • 2.1 cm'/sec. 
Axial velocity * 291.2 cm/sec. 
Zero-g minimum length * 2.1 cm, 
Steady-state zero-g length « 3.8 cm. 
Steady-state zero-g maxknum radus * 0.72 cm, 
Reynolds no. ■ 783. 



Steady state zero-gravity jet diffusion flame 
(Cochran, 1972) 



- Science Appfcabons Mamaftbna/ Corporation - 



2.2-SEC. DROP-TOWER EXPERIMENTS /CONT.J 



mm 









it 


1 BBS- 

-Si- 

_is§lfe 



^J*! 

£* 



fe-V 



Ptrlg 

■-1 K^a;. 




Nozzle radius =0.186 cm, 
Normahg length r 2.78 cm, 
Methane flow rate = 2.8 cm'/sec. 
Flow velocity « 25.44 cm/sec, 
Zero-g minimum length = 1.86 cm. 
Extinguishment length = 2.93 cm. ' 
Reynokfc no. • 284 



Ti««la nnitnitf. 



Extinguished zero-gravity jet diffusion flame 
(Cochran and Masica, 1970) 




| — Sciano* Applicators MwnMfenaf Corporate!- 



2.2-SEC DROP-TOWER EXPERIMENTS (COM) 



3.5 



53.0 



Stttdy - 
Jttte, — #• ".' 

lero-g / 

length . 



'•••••••• 



(b) 



Z.5 



Z.O 



b>y 



y, *»t. noraal-g 

r- *• «f» length 

' _Mn. ztro-g length 



Tint, s«C 

Cochran (1972) 

Methane flow rate < 2.1 cm/sec 
Nozzle radto > 0.051 cm 
Nozzle velocity « 291.2 cm/sec 
feynoida no. • 79j5 



7.9 


] 


(•) 






' Exflnctbn 

i >• 


■ 

c 

V 


■ 






• 


! • 


!«•» 


- 








i' • 


IT 




1 


JL 




! Ztro-q 


«. 


s 


4 


» 


Tlat 


5.3 5.7 
SK 



Cochran and Masica (1970) 

Methane flow rata ■ 2.8 cm7aec 
Nozzle radkia > 0.186 cm 
Nozzle velocity » 25.44 cm/sec 
neynolda no. ■ 284 



. SdencaApplicaiionati t aiiiat K irtalCcrpont i on- 



WOP-TOWER RESULTS 



20 



16 



12 



(COMPILED FROM COCHRAN, 1972) 



A STEADY STATE 
■ TRANSIENT 
• EXTINGUISHED 



NOZZLE RADIUS, r Q - 0.0825 cm 



' ■^■"" 



r Q - 0.051 cm 




DATA AVAILABLE 
NO DATA AVAILABLE 
I ■ ' i 1 1 1 1 1 1 — 



20 40 60 80 100 120 140 160 180 200 

REYNOLDS NUMBER 



, — Sb«r»AB*afens*tfarn«anar Corporation- 



THEORETICAL ANALYSES 



• STEADY FLOW 

• TRANSIENT 

• s-JITTER ANALYSIS (DISCUSSED UNDER 
•RESEARCH REQUIREMENTS" SECTION) 



■ — SamK»Appto*tkxvHwn*k^ Corporation- 



STEADY HOW AMI YS/S 



rHUIMUM TWPUMUK iuUUX 
> rf LUX fHOMT") 



fOKNIMl COU HOSN-v t 



JL 



miiial 'mi 
nuocirr nviu 

•i 

irffe}- 

WHIM MASS 'UCTION 

oismtuTim 

lit? 

mum hm*imiun 
oishiiuiim 






c-iouNOtursorTic 

)\D»FUSICNaiCION 



-V- 









It' 






oucsaNT 

WVWCWWUfT 

f . 
T, 

FUU. •(, MASS 

i, riuciiOKSi 

•,«,- iroiniiKFKi 

■UCTIOMI 



lif of vtrlicfcl Umintf j«t. 



V 1 " ,, • vIV*' 






met u*n 



w(i 



if J M »*» 



WHICH 




»0-»c-}«H «0-«,-}«H 

sncHOMCiic iou««UNai iaib 

Bck«m«li« of "compteu oombulbon" «quUitanuin ehomulry modal. 



I 



STEADY, AXISYHHETR1C FLOW 

2-D CONSERVATION EQUATIONS FOR HASS. SPECIES. HOHENTUH. AND ENERGY 

PARABOLIC FORH 

WULTICOAPONENT DIFFUSION 

INFINITELY FAST REACTIONS 

RADIATION 



sd 



-Soenc*Applicatk>nitoarr*ion4 Corporation- 



STEADY HOW AHALYSIS (COHT.J 



Continuity 



Momentum 



8 (our) ^ a(pvr) 



-- du -- 8u 
pu j= ♦ 



7 V 1h . ,X_, I HurlZ/ir) 
P 8r W ? dr 



+ <£?' 25 



Diffusion of Elements 



— Jo* — jcT' 

pu r=- + pv ■*=■ - 

ax dr 



Re r dr 



Energy 



pu ax * pv ar W ? ai 1 Pr a? " r = (h * »}♦ V 



-Sofcnce 



Applications Memefional Ccwxwtfor? 



^toW nOW ANALYSIS (COHT.) 



P " RT £<*Vw 
1 



. i . all gaseous species 



h -r 

X 



aV 



where 



H - u /2 



P -P* 

Gr ■ ^ l p* ' 



p»u»r« 

«• ■ — ^r - 



u 


■ 


u/u« 


r 


■ 


r/r< 


P 


■ 


p/p ( 


H 


■ 


H/H 


V 


■ 


v/u 



x ■ x/r* 
ji ■ v-'t* 

P e -P p e'° 
*P " p e -l " P e "P* 



■V 



l _ »a ~t _ kt w X 

i- u ■=?- a " E " ~~v a 



X w 



pu |r- + Pv Jf " ReSc r dr ir " 3r 



and 



— aS -- »3 . i- i *_ { St I" 1 - ^ + e 



r h 



If]} 



♦ <L 



-JCHT©^ 



r 



ScMnc«App/»c««o«Wwn«*x>»/CW)<^ 



'sJiffirnMlMtTsislcoNT.) 



Boundary Conditions 

« x - 
f H-H o (r 
< r * r^ / u -u(r 



Note: Arbitrary initial 
profiles nay be specified 
including parabolic for 
the velocity. 



r > r , 



»1 ■ a io ( 
v - 



H - H 



r) - 1 for pure fuel 



u - u e (u e " ° for 



the quiescent environment 



a i " a ie 

9X30 I 



u - 
H-H # 

•i - a i. 



00. 
»u „ *H » — i. - (symmetry ) 
B r or »r 



where 



— Scimc* Applications himatianalCorporlion- 



STEADY FLOW ANALYSIS fCONT.J 

Multicomponent Diffusi on 

i" 1 J kj j-i w j J 8 ^ 8r ' 



E j J - 
J - 1 



i l - T * U 4 i* 



t . .2-663 x 10~ 2 t 1 - 5 



kj 



r-n-' 



(1.1) 
kj 



' kj sec' 



Where [ T ] - ° K , and r P i - $- 

J m * 

In addition. 






with th. collision cross-section given by: 

O 4 (J 



kj 2 



^ 



(1,1) 



1.5146 - 0.62499 T + 0.10023 T 2 



\i - * n <"v 



'kj ■ <VJ < K > 



" So * nw ^«>*e««afB*i»fn««on i /Q»por«ifc»i- 



TO^ «W ANALYSIS (CONT.) 



Thermal Conduct Ivlty 

The Mixture rule* for the thermal conductivity are given by: 



K gases 

X - Z 
k-1 



i£ 



V£ 



. 1 



N ^ h j> 



j-1 k * ** j-1 

jrtc jrtc 



Joule 
(m-sec-°k) 



J 



** - 3 - 75 2h 



\"« »*-*** 



[0.352 C* W* - o.88] 

" * ' " f * " 8.3143 x 10 3 — tel °v 

ft kgm-mole 






"kj 



_ L 1 +/£ *i 1 



r kj 



Not* that 






k w 



k * 



. Science Applications 



Memationa/ Corporation 



STEADY now ANALYSIS (CONT.) 



Viscosity 



n gj 


ues 


<*v"* 


jj " E 
where 


jA 

w «, 



r *j 






with 



and. 



-6 yL,* 



2.6693 x 10 -/TW ,_JS3 — ) 



!< 2 ' 2) - 1.6507 - 0.6688 T + 0.L072S T k » 



Science Applications International Conxnton 



MWKHPREVCTIOMS VS. EXPERIMEHTAL RESULTS 



Km 



BOFUUMXNk ftttic* 
* • Sc - I 

ttn* mwoktonsiw 

"• CMI / //. 

1 01 / / /' 

-- » //// 



.- 




»>utrnc wjoeiKxis 

r,ICMI fr K 

ion ti i« 

.0»l .1 » 

o tmnwwi MI * * 



CMI 



J I I I I I I I 

q, cmVkci 




, -i t • ■ a m 

0, cm'/jici 

1.«<mm< •< Schmidt Nu»bw «»••> vuitioi f-HtUm, of mcimI- 
|r*«Ujr Hum Iwfth. 



* Reproducedfrom Edehian et al (1973). 



I — Sdtnc* Applications toematim* Corporati on - 



AHALVTICAL PREPICTIONS VS. EXPERIMEHTAL RESULTS (CONT.) 



too ^ 




w 

II 
I 



Co«>l, t« 



■ ■ ■> 




-I 1— LXi ■ ' i 



•«. 




froitu rim, dl/tlr t«l/ 
.— '•II caaalltr; aMtl 



■j« MX CO Ml; 

/\f /. »,0 •* C(l) ml/ 

<o 

J — I — I ' ■ I 

••« 0.1 I.J I.I j.o J.« M 



EipeHmentil data*'* 


Nonml -g 


r, (cm) 


O 


0.051 


A 


0.0B2S 


O 


0.113 





0.186 





0.318 


V 


0.442 



Solid synbols denote 
zero .9 



Partial oxidation 
theory 



H,0. CO H-0, C($) 

O t> 

A 



0.0S1 

0.082S 

Solid sjaiboH denote 
experimental djti* 



r, (oa) 



Oitl* 



Q ■ 


1.08 em 


/«ec 


Re,. 


40.9 




\' 


294 K 




t ■ 


1 ItM 




r . • 


0.0S1 ca 





(Reproduced from Edeknan et al 1973) 



•Cochran (1972) 

' Cochran and Mastca(1970) 



— Sem^AivHcatknhlarTmlionalCovormtion- 



wm ■" " ™ — ■■"»« w ohvjt* itajl*ut7eaW ViTja^ 1 *r7TeWrlJa^ — H 

'"""'"'WfHMI^WM/trAt RESULTS (COHt, 

EFFECTS OF NEGATIVE-G (INVERTED FLAME) 




feat fit* 'IU (oaVtat) 




(k) ■ lm-noo 1 

□ 1S00.I700 I 
■ I100-IH0 I 

Bi 



0.1 
in 
l.i 

I... 



MOO 


(c) 

• • 

>' ":\ 

••' :•'••• 

« • • •* « 

V.UL 


•artlcil 
latitlaa (U) 


taoo 

IMO 


• 0.1 

• *J 

» I.J 

• I.I 

• 1.0 



Mill alltaac. (la) 



-1 i — • 

•Mill Kltwn (l«» 



<*) 


'artlcil 
lacatlaa |l.| 


■ l.i 


■ 


• 0.F 

* I.I 

-/»— • I.I 


• .• .•• • 
••• .\ 





.; 



• «.J o.« II |.| 

•Mlal altuaca (la) 



Methane Injection velocity » 0.17 - ^2 H/aec In Fig. (a) where 
characteristic* of normal-g (A), rero-g (B). negattve-o (C) 
stagnation points (D). recirculation <E). and Kt-eff (F) are shown, 

« « ? "^"J - ** * "•* ***«* • e-4 nVsee, fueHube ahmeter* 
0.19 In, and duct dameter « 6 h. ««nner 



Fuel Injection velocity . 1.2 ft/sec h (b). (c) and (d). 



- Science Applications International Corporation - 



PHOTOGRAPH OF THE INVERTED FLAME 







, Science Applications krtemational Corporation - 



(a) t<0 



TRANSIENT ANALYSIS 




inert- 

oxidizer 

boundary 



i ' 



r f U,t), 



Inert- 
fuel 
boundary 



(b) t>0 

• CONSTANT PRESSURE 

• LEWIS NUMBER OF UNITY 

• EQUAL AND CONSTANT FLOW VELOCITIES, 
TRANSPORT PARAMETERS. AND PHYSICAL 
PROPERTIES 

• FLAME SHEET APPROXIMATION 

• NO BUOYANCY EFFECTS 

• BOTH AXIAL AND RADIAL DIFFUSION INCLUDED 



"k 



aflame 
boundary 
at time 
t 2 Ot,) 

, flame 
boundary 
at time 



d72 



- Science Applications Mamrtnnar Corporation - 



TRANSIENT ANALYSIS (CO NT.) 



GOVERNING EQUATION: 



_, 














12* 

at 


r iz 


*G 


ar 


+ ar 1 W/ . 












»ir<r f 


' z>o, 


* >o - 


zt 


u2!a. 

31 


*(+ 


22k 


ar 1 


9\\ 




k 








r<r<j'/», z>o. 


t>«, 



o<r<J7* * z>«, t>o . 

WHEJU U • U . U - CONST. , D - D_ m D - CONST. , 



Y(r.«,t). 



Y F o<r<r f , z>o,t>o, 
-vr r<r<wy»» z>o, t>o, 



y - (GRAM OF F/fiMH OF <>)„,..„ 



i — Soonc* AppBcations ht^m^^^o^^r^ Corpora ti on - 



TRANSIENT ANALYSIS (CONT.) 



INTTIAL CONDmONS: 



(0 o<r<J/i , z>0, t-o, 

~ yY o,. J/»<r<J'/2, z>o, t*o, 

Y{r,o, *>-{*" °^ r<J/i ' Z -°' t>0 ' 

l-VY 0# . At\<.r<i'/x. z-o, f>o. 



SYMMETRY CONPmON: 

aY(o,z,t)/9r - . 

BOUNDARY CONDITION: 

*Y(«7a, z, t)/ar - . 

FLAME LOCATION: 

Y(iJ. z,t).o. 



-Sc«ne»Appfic*6ons*i«OTrtionrfG)rpof*ion 



TUNSIEHT ANAMIS (CONT.) 



TRANSENT FLAME SHAPE: _ 

r. >(Y r ..->VY ,.)(-f)J ^jji^^L 

2- »jUW*j*) 



J-i 



x exp 



<*+h) 



aDt 



1-11 + 



^tKt + ^i^^ 



" i (t-§-r 



-l(\,-^o,)(7f-^o.J 



* exp • 



«(*♦&) 



aot 



'1 



4 D t^fl-^^Sd 



WHERE Aj(J>0 *«™E ROOTS OF J, fa *'/*) - • 

RTF ADV-STATE FLAME SHA PE f*-^") 



-^^^^cMpSrFson between transient analysis 

and prop -tower results 



^__« Ockria •"* RMtca (1») 

Tr.ml.M. tlliftlc haljrtlul —*t\ (II) 

Tlmrttlctl SUUj SfU in) 





J«_". 4 _°i 



Di binary diffusion coefflcent 
of methane In nitrogen. 



The vilue of ■ 1.0 en /»te 
corresponds to T • 750 K. 



I Scienci Applications International Corporation - 



0.0 



COMPARISON BETWEEN TRANSIENT 
ANALYSIS AND DROP-TOWER RESULTS 



-»--'^°2. 



3.24 




cm/ sec 

tea, 1970) 
Transient, elliptic analytical model 
Theoretical steady state 



0.5 



1.0 




1.5 
Time, sec 

• 0.80 cm '/sec 



2.0 



2.5 



. 0.85 



7 -J1- --7 

J. 94 



1.00- 



3.35 



*,« 0-051 an, U • 291.2 cm/sec 

— Experiment (Cochran, 1972) 

- Transient, elliptic analytical model 
Theoretical steady state 

1 — ■ — ' — <■ — i — i — ■ ■ ■ ■ ■ i 



0.5 



1.0 l.S 
Time, sec 



2.0 



2.5 



I — SoerKaApplkatkmkUemationalCorporatk 



Da binary diffusion coefflcent 
of methane In nitrogen. 

The volues of D=1.0 and 0.85 
cm2/sec correspond to T=750 
and 650k, respectively. 



EFFECT OF AXIAL DIFFUSION 
(TRANSIENT ANALYSIS/ 




vim turn oirrujio" 

WIlHOUT klMH. DIUUSIM 



R ?°- 318 1.0 



o • s.o »'/•« (c \/H % ot IV* K) 

U • 18 nfifl 



« ■ Jli u 



-I " ~ 

CM Jt\, 



I 
2.0 



SdenceAppfcatmraWernaf tana/ Corporation 



BACKGROUND SUMMARY 



i 



LAMINAR GAS JET DIFFUSION FLAMES : 

• BUOYANCY SIGNIFICANTLY ALTERS FLAHE STRUCTURE 

CHEMICAL KINETICS-RADIATION-TRANSIENT INTERACTIONS 
ARE IMPORTANT UNDER REDUCED-GRAVITY ENVIRONMENT 
ADDITIONAL MICROGRAVITY DATA IS REQUIRED TO 
OBTAIN SUFFICIENT QUANTITATIVE UNDERSTANDING OF 
LAMINAR GAS JET DIFFUSION FLAMES 



. SoenceApplicationstiternationalCaporahon- 



PROPOSED PROGRAM 



r— Scmnca Applications tMmafJanaf Corporation - 



PROGRAM OBJECTIVES 



• OVERALL: 

GAIN A BETTER FUNDAMENTAL UNDERSTANDING 
OF LAMINAR GAS-JET DIFFUSION FLAMES IN 
GENERAL. AND EFFECT OF BUOYANCY IN PARTICULAR. 

• SPECIFIC: 

OBTAIN NEW DATA FOR ELIMINATION OF INADEQUACIES 
IN THE EXISTING DATA AND THEORY. THIS WILL BE 
ACCOMPLISHED BY STUDYING: 

• UNSTEADY PHENOMENA OF IGNITION 
AND FLAME DEVELOPMENT 

• STEADY-STATE FLAME STRUCTURE 

• RADIATION AND KINETICS 

• QUENCHING PHENOMENA 




I — ScMnosAflp^fonshfematJonaf Corporation- 



SUMMARY OF 
RESEARCH REQUIREMENTS 



The research requirements are detailed h the Science Requirements Document 



Science Applications International Corporation 



TOTAL PROGRAM AND MINIMUM MATRIX OF TESTS 



OPERATING CONDI TONS 
AND RANGES 



NUMBER OF EXPERIMENTS 
FOR TOTAL PROGRAM 



FUEL FLOW RATE ■■ <15 cm 5 /sec 



NOZZLE RADIUS = 0.05-0.15 ch 



FUEL: METHANE, PROPANE 



CHAMBER PRESSURE - 1/2 AND 1 at* 



CHAMBER OXIDIZER: AIR (0 2 /N 2 - 
0.21/0.79). AND AN OXYGEN - 
DEFICIENT 2 /H 2 MIXTURE (0 2 /N 2 
0.10/0.90) 



TOTAL NO. OF FLAMES: 



18 



MINIMUM MATRIX 
OF TESTS 



1 (METHANE) 



1 (AIR) 



12 



Science Applications htemational Corporation- 



RATIONALE FOR TEST MATRIX SELECTION 

COMBINATION OF THE EFFECTS OF INJECTION VELOCITY AND NOZZLE SIZE ON FLAK 

CHARACTERISTICS HELP SORT OUT THE EFFECTS OF DIFFUSION PROCESSES. 

EFFECT OF REYNOLDS NUMBER ON STEADY-STATE. TRANSIENT AND QUENCHING PHENOMENA 

CAN BE STUDIED. 

EFFECT OF FUEL REACTIVITY. STOICHIOMETRY AND DIFFUSIVITY ON FLAME STRUCTURE 

AND BEHAVIOR CAN BE STUDIED USING DIFFERENT FUELS. 

CH„ IS SELECTED BECAUSE OF THE BASE OF DATA THAT ALREADY EXISTS. CjHg IS 

SELECTED TO EXAMINE A HEAV1ER-THAN-AIR FUEL. AND BECAUSE IT IS REPRESENTATIVE 

OF A WIDE RANGE OF HIGHER-MOLECULAR-HEIGHT HYDROCARBONS. 

EFFECTS OF PRESSURE ON SOOT FORMATION. QUENCHING. RADIATION. KINETICS. AND 

IGNITION CAN BE STUDIED. REDUCED PRESSURE IS USED TO ENHANCE KINETIC EFFECTS. 

AMBIENT COMPOSITIONAL EFFECTS ON FLAME STRUCTURE. KINETICS. SOOT FORMATION. 

AND IGNITION CAN BE STUDIED USING DIFFERENT CHAHBER OXIDIZERS. REDUCED LEVELS 

OF 2 ARE SELECTED TO ENHANCE KINETICS. 



, Sconce Applications International Corpora t ion - 



RESEARCH REQUIREMENTS 



FUEL 



EXPERIMENT TIME 



FUEL TANK 



EXPERIMENTAL CHAMBER 



INITIAL CHAMBER PRESSURE 



CHAMBER ENVIRONMENT 



PURE. HIGH-QUALITY METHANE; PROPANE HILL BE CONSIDERED 
FOR LATER FLIGHTS; METHANE VOLUME FLOW RATE = 0-15cm 3 /sec 
PROPANE VOLUME FLOW RATE = 0-5cm 3 /sec. 



UP TO 10 SEC FOR EACH FLAME. 



INTERNAL PRESSURE = TBD (BASED ON THE NUMBER AND 
CHARACTERISTICS OF FLAMES STUDIED DURING EACH TEST); 
5-15 ATM . 



EITHER CYLINDRICAL OR RECTANGULAR; HIN.HE1 GHT/DI A. (OR WIDTH) 
- 1.5. VOLUME (IF IN MID-DECK) = 0.01 h 3 (MUST FIT INTO TWO 
LOCKERS); VOLUME (IF IN MSL ) > 0.01 m 3 (TBD BASED ON THE 
NUMBER OF FLAMES PER FLIGHT; LESS SEVERE RESTRICTION ON 
OCCUPIED SPACE COMPARED TO MID-DECK); VOLUME (IF IN 
SPACELAB) > 0.01 m 3 . 



1/2 ATM AND 1 ATM. 



STD. AIR (21Z 2 ,79Z « 2 ). ALSO. 10Z 7 . 90Z l» 9 FOR 
LATER FLIGHTS. l 



I — SciarKaApplkxtkrulntamationel Corporation 



RESEARCH REQUIREMENTS (CONT.J 



0, CONSUMED 



NOZZLES 



FLOW CONTROL ACCURACY 



ACCELERATION ENVIRONMENT 



OPERATOR INVOLVEMENT 



TOLERANCES 



slOZ AT THE END OF THE FLIGHT. PREFERABLY 1Z 
3Z AT THE END OF EACH FLAME MEASUREMENT. 



RADIUS - 0.05 - 0.15 cm. LENGTH =15 cm. 



- 0.01 cmVsec. 



<10" 3 6. 



t FLAME IGNITION 

• MOTION PICTURE PHOTOGRAPHY OF FUME 

• VISUAL OBSERVATION 

• TEMPERATURE. PRESSURE AND ACCELERATION MEASUREMENTS 

• SAMPLE REMOVAL 

• RADIATION MEASUREMENTS. 



SAS AT BEFORE EACH TEST - ± 5K (2D; 

GAS AP BEFORE EACH TEST - ♦ 0.3 PSIA (21); 

A0 2 AT THE BEGINNING OF THE TEST SET - ± O.ImoleZ. 



So«nc»Appfcrt ra **arT*iond Corporation- 



RESEARCH REQUIREMENTS (CONT.) 



IGNITION 



CAMERA FRAMING RATE 



VIEWING (RESOLUTION) 



DATA RECORDING RATES 



EITHER SPARK OR HEATED WIRE (SPARK PREFERRED); POWER 
SUPPLIED BY SHUTTLE: FINAL DETAILS OF SUBSYSTEM TBD 
FROM THE 5-sec DROP-TOWER TESTS. 



HOO FRAMES/SEC FOR THE FIRST 5 SEC. OF MEASUREMENTS 

(TO STUDY IGNITION) AND 100 FRAMES/SEC FOR THE REST OF THE 

PERIOD; i.e.. TWO-SPEED CAMERA IS PREFERRED. 



FIELD OF VIEW MUST BE 5-ch HORIZONTAL BY 10-cm VERTICAL 
AT THE PLANE PASSING THROUGH THE FLAME CENTERLINE AND 
PERPENDICULAR TO THE AXIS OF THE CAMERA. 



THERMOCOUPLE SAMPLING RATE - 20/SEC AND 
TRANSDUCER SAMPLING RATE - 20/SEC (PREFERRED). 



, Somv»Aeel**tkrat*am#xrmlCaparmtian- 



ACCELERATION 



i 



ACCELERATIONS GREATER THAN 10- 3 6 RESULT IN (a) INCREASE IN VELOCITY OF THE 
BURNED GAS AFTER REACHING A MINIMUM AT DOWNSTREAH (TAKE-OVER OF CONVECTIVE 
EFFECTS). AND (b) SLOWER DECAY OF GAS TEMPERATURE DOWNSTREAM OF THE NOZZLE. 
BURNING WILL BE AFFECTED BY G-LEVEL. BUT MEASUREMENTS WILL NOT BE ADVERSELY 
AFFECTED BY VARIOUS 6-LEVELS. 
SEE THE FOLLOWING ANALYSIS FOR THE EFFECT OF RESIDUAL ACCELERATIONS U-JITTER). 



EFFECTS OF GRAWTATIOHAL ACCELERATION 













»■'• 








1 y'i 








^ y?% *»/% 


-£- 






C.t 0.4 0.1 

in. 


1 ill 


=fr- 



a 




•.I l« a.i •• ■ • 



X/L 



U s Jot velocity, 93.5 cm/sec 

Nozzle radkjs = 0.0825 cm 

L > 50 cm 

T max * maximum temperature reached at the centerfna 

T » Initial gas temperature, 300K 



-ScmncmApplicatiorahlomalionalCaporwIion- 



CONTNUITY : 



EFFECT OF RESIWAL ACCELERATIONS 

ORDER-OF-MAGNITUDE ANALYSIS 



(l) 



ASSUMPTIONS: f m ft , «.,?♦$ e'"* * m ?+$j>\ 

SUBSTITTUION IN THE CONTINUITY EQUATION YIELDS THE FOLLOWING TWO EQUATIONS FOR THE 
STEADY-STATE AND TIME-DEPENDENT COMPONENTS: 

USING [£] = -J-, [^-Js-J- AND ["f~] H + ' WHERE L " FLAME LENGTH AND R " MAXIMUM 
FLAME RADIUS, WE THEN OBTAIN 



7? 3L 



A. A 

"R 3L 



(4) 



.ScienaA&KatkimkrlemaiionalCoiponlion- 



EFFECT Of RESIDUAL ACCELERATIONS (CONTINUED) 



ASSUMPTIONS: 



(5) 



f .f, t .t+l. u * »-^< ««♦*•'•'*' 



C0» S ,«»T WSCOSITY. »ND «UG,BLE H.G.E.-O.DE. POTUMnO. TO*. I-C 

, !, " t TERMS. 

SUBSTITUTION IH THE IWIEKTW EOUftTlOK YIELDS 

(7) YIELD, RESPECTIVELY 



(6) 



(7) 



) AND 



«~SK-/^8F-]< 



i»i 



S3 






(8) 



WHERE WE HAVE USED EQS. <«.! * IS THE KINEMATIC VISCOSITY. ^IPiKS 



. Scimx*Acclk*6omlr*mMK**ICa«>ai*ion 



EFFECT OF RES/DUAL ACCELERATIONS /CONTINUED; 



ENERGY: 

ASSUMPTIONS: 



/-/^ 



luit 



_ a. it»" M a i 

u»u+ue , v«v+ve 



'-* 4-^f. 8- ! 



■ ,MF PD - X/ Co - CONST.), NEGLIGIBLE KINETIC ENERGY, AND 
EGU I E SECOND-OR PERTURBATION TERMS. THE ASSUMPTION OF Le - 1 
G COUPLES E THE ENErSy-AND ELEMENT-CONSERVATION EQUATIONS. 

THE FOLLOWING EQUATIONS WE THEN OBTAINED FROM EQ. (9): 

A M A C* 



EQS. (10) AND (ID YIELD .. g , ; 1 



(9) 

(10) 

(11) 



Kl 






(12) 



WHERE WE HAVE USED EQS. (11. SUBST ITUTION OF EQ. (8) IN (12) YIELDS,, 



(13) 



Scone* Applications International Corporation- 



EFFECT OF RESIDUAL ACCELERATIONS /CONTJ 

CALCULATIONS FOR FLAME A l ' 

CH„/AIR FLAME, u =291.2 cm/s, r Q =0.051 cm. NUMERICAL SOLUTION BASED ON Le=Pr=l HAVE SHOWN 

THAT FOR 0-G: L f -0.209 ft-6.37 on, W 0.033 ft=l an, AND T = 1620 K. SINCE Scl, D= ,.<,. C 

cm'/sec AT T=1620K FOR DIFFUSION OF CH„ IN AIR. THE FOLLOWING PLOT OF h vs. L f FOR DIFFERENT 

RADIATION COEFFICIENTS (RF, SHOWS THAT Ah=2.3*loV f (WITHAL, IN ft,,. ALSO. NUMERICAL RESULTS 

SHOW THE TAKE-OVER OF CONVECTIVE EFFECTS FOR ACCELERATIONS GREATER THAN 10" 3 G (SEE BELOW). AT 

10- G. L f =0.206 ft-6.28 cm. WE DO NOT ALLOW MORE THAN 33* VARIATION IN DIFFERENT INFlflMF unr.urc 



BETWEEN 0-G AND 10" 3 G. THEREFORE Ah=(2 . 3x10^x0. 33(0 .209-0.206,=23. NUMERICAL RESULTS SHOW THAT 
h=550 AT THE FLAME TIP. HENCE |fi|/h=23/550=0.042. 



H=50 cm 
r o =0.051 cm 
u =291.2 cm/s 




r o =0.051 cm 
u =291.2 cm/s 
Le=Pr=i 



0.2 



0.4 0.6 0.8 
x/H- 




S -800 



1.0 



0.19 



J Ah " 2.3xlpy f j 



0.20 



0.21 0.22 

L f (ft>- 



0.23 



— SdunctApolkMionskitmrnmtinnrtnnr^-s^ 

EFFECT OF RESIWAL ACCELERATIONS /CONTINUED) *~ 

CALCULATIONS FOR FLAME A (CONTD) 

SUBSTITUTION OF L f »6.37 cm, R^l cm , fr 1/2 M5 ^ ^^ ^ 

( P e /P» = (T/T e )= 15 20/300= 5 ., (ASSUMING AIR,, AND° |fl/ff . 0.0,2 IN THE EXPRESS IOnV^S 

g/g - 0.000185 ^/(w 2 ♦ o.uxw 2 + 51.5,. 

THE FOLLOWING DIAGRAM SHOWS THE ALLOWABLE FREQUENCIES. 



10 



,-1 



t 10 

o 
01 

<oi io" 3 
10"" 



2 - 



u -291.2 cm/s 

r -0.051 cm 

L f -0.209 ft - 6.37 cm 

(AL) allowedAf0.5X 

THRESHOLD FOR TAKEOVER 
OF CONVECTIVE EFFECTS 



10 - 




-6 



-«t 



-2 -1 
LOG w — 



[DAMPING IS REQUIRED 
-L 1 , 



*2 



, SciencaA&k*tiomtotam*ianalC<xpor*ion 

EFFECT OF RESIDUAL ACCELERATIONS (COM.) 

CALCULATIONS FOR FLAME B 

CyAIR FLAME, U Q =93,5 cm/sec, r o «0.0825 cm. NUMERICAL SOLUTIONS BASED ON Le=Pr=l HAVE SHOWN 
THAT FOR 0-G: L f =0.191 ft-5.82 cm, ^-0.01 ft-1.22 cm, AND T fla(ne tlp =H20K. SINCE Sc=L 
D=,=2.0 crn^s AT T-1120 K FOR DIFFUSION OF l\ IN AIR. THE FOLLOWING PLOT OF h vs. L f FOR 
DIFFERENT RADIATION COEFFICIENTS (RF) SHOWS THAT Ah-lxio" AL f (WITH AL f IN ft ). ALSO, NUMERICAL 
RESULTS SHOW THE TAKE-OVER OF CONVECTIVE EFFECTS FOR ACCELERATIONS GREATER THAN 10" 3 G (SEE BELOW! 
AT 10' 3 G L f -0 173 ft=5.27 cm. WE DO NOT ALLOW MORE THAN VSI VARIATION IN DIFFERENCE IN FLAME 
HElGHIi BETWEEN 0-G AND lO'V THEREFORE Ah=(lxl0 4 )x0.33x(0.191-0.173)=60.NUMERICAL RESULTS 
SHOW THAT h=850 AT THE FLAME TIP. HENCE |fi|/h = 60/850-0.07. 




c. 



-100 ■ 
-300 



| -500 



-700 
-900 




r Q =0.0825 cm 
u =93.5 an/s 



t 5 Le-Pr-1 
Ah=lxlO H AL f j v^ 2 

4-"-vr ' 

• ■ ■ i . 1 1 . . . 1 1 ■ . — >_ 



0.13 



0.17 



x/H 



0.21 
L f (ft)- 



0.25 



, — Scm& Applications HemationalCapOMtion- 



EFFECT OF RESIDUAL ACCELERATIONS (CONT.) 

CALCULATIONS FOR FLAME B (CONTD) 

SUBSTITUTION OF L f - 5.82 cm, u - 1/2 u Q = 45 cm/sec, R^ - 1.22 cm, 
D - »• 2.0 cm 2 /sec, g - 980 cm/sec 2 , IP^/P) - (T/T e ) - 1120/300 - 3.73 
(ASSUMING AIR), AND | h | /h - 0.07 IN THE E XPRESSION FOR ?/g Q GIVES 
•Syg - 0.000«6V(<J 2 ♦ 0.01) fw 2 + 6.09). 

THE FOLLOWING DIAGRAM SHOWS THE ALLOWABLE FREQUENCIES. 



10 



-1 - 



u -93.5 cm/s 
r -0.0825 cm 
L f -0.191 ft - 5.82 cm 

^allowed / L f " 3X 



Thresold for takeover 





- Sciancv Applications titemalional Corporation - 



DATA ACQUISITION 



RECORD FLAME IGNITION AND DEVELOPMENT DURING COMBUSTION 
RECORD GAS TEMPERATURES AT DIFFERENT LOCATIONS ABOVE THE FLAME 
CONTINUOUSLY RECORD CHAMBER PRESSURE AND TEMPERATURE 
RECORD ACCELERATION" 
GAS-SAMPLE REMOVAL 
RADIATION MEASUREMENTS 



r 



Sdanc* Applications International Corporatk 



MEASUREMENT TECHNIQUES AND APPLICATIONS 
TO MODEL DEVELOPMENT 



MEASUREMENTS 



CINEMATOGRAPHY 
(USING HIGH-SPEED 
MOVIE CAMERAS) 



RATIONALE/APPLICATIONS TO MODEL DEVELOPMENT 



THERMOCOUPLES 
(USING RAKES) 



SAMPLING PROBES 
(USING SAMPLING 
BOTTLES) 



PRESSURE 
TRANSDUCER 



OBTAINING TIME-RESOLVED VISUALIZATION DATA ON 
IGNITION. FLAME DEVELOPMENT. COLOR. LUMINOSITY. 
AND QUENCHING. FLAME SHAPES CAN BE COMPARED 
DIRECTLY WITH MODEL PREDICTIONS. 



OBTAINING TEMPERATURE DISTRIBUTIONS IN THE FLOW 
FIELD ABOVE THE FLAME. THESE TEMPERATURE DISTRI- 
BUTIONS CAN BE COMPARED DIRECTLY WITH THE MODEL 
PREDICTIONS OF THERMAL FIELD. 



OBTAINING SPECIES DISTRIBUTION ABOVE THE FLAME 
WHICH CAN BE COMPARED WITH THE MODEL PREDICTIONS 
OF SPECIES FIELD. 



MEASUREMENT OF THE INCREASE IN CHAMBER PRESSURE 
DUE TO THE BURNING PROCESS YIELDS ESTIMATES OF 
THE OVERALL HEAT RELEASE RATE. 



TOTAL 
PROGRAM 



YES 



YES 



YES 



YES 



MINIMUM 

MATRIX OF 

TEST 



YES 



YES 



YES 



YES 



So^KaAppftcaforoWam^JonWCapewltan- 



MEASUREMENT TECHNIQUES AND APPLICATIONS 
TO MODEL DEVELOPMENT (CO NT.) 



MEASUREMENTS 



VELOCITY 
MEASUREMENT 



RATIONALE/APPLICATIONS TO MODEL DEVELOPMENT 



RADIOMETRY (TBD) 



MEASUREMENTS FOR THE SCALE OF THIS EXPERIMENT 
REQUIRE A LASER-BASED DIAGNOSTICS. WHICH IS NOT 
FEASIBLE AT THIS TIME. THIS REQUIRES A COMPROMISE 
IN WHICH PREDICTIONS AND OTHER MEASUREMENTS WILL 
BE USED TO PROVIDE THE VELOCITY. 



TOTAL 
PROGRAM 



NO 



ISOLATION OF THERMAL RADIATION. AND AS A RESULT. 
DETERMINATION OF THE EFFECT OF RADIATION ON 
REDUCTION IN TEMPERATURE WITHIN THE FLAME. 
POTENTIAL EXISTS FOR DETERMINATION OF SOOT CON- 
CENTRATION AND GENERATION. EVIDENCE SHOWS THAT 
RADIATION FROM SOOT IS NOT DOMINATED BY C0 2 AND 
HoO. THEREFORE. CALCULATIONS FOR C0 2 AND H 2 
COMBINED WITH MEASURED RADIATION YIELD THE CON- 
TRIBUTION FROM SOOT. AND HENCE. SOOT CONCENT- 
RATION. 



MINIMUM 
MATRIX OF 
TO 1 
NO 



YES 



YES* 



•IF EARTHBOUND EXPERIMENTS SHOW THAT RADIATION MEASUREMENTS ARE REDUNDANT COMBINATION 
OF THEORETICAL ANALYSES AND EARTHBOUND RESULTS WILL BE USED TO DELETE THOSE MEASUREMENTS 
FROM THE SPACE PROGRAM. 



r 



Sconce Applications Mam««nafCarpor*>on - 



MEASUREMENT TECHNIQUES AND APPLICATIONS 
TO MODEL DEVELOPMENT (CONT.) 



MEASUREMENTS 



RATIONALE/APPLICATIONS TO MODEL DEVELOPMENT 



ACCELEROMETER 
(IF AVAILABLE) | 



MEASUREMENTS OF THE GRAVITATIONAL ACCELERATION 
ENABLES US TO INTERPRET THE DATA THAT COULD 
BE AFFECTED BY GRAVITATIONAL LEVELS GREATER 
THAN 10" 3 6. 



TOTAL 
PROGRAM 



YES 



MINIMUM 

MATRIX OF 

TESTS 



YES 
(IF AVAIL- 
ABLE) 



^SciencmApplicMtionshtomationalCorpontk 



RATIONALE FOR OBSERVING THE COLOR OF FLAMES 

THE WOIFHARD-PARKER SLOT-BURNER HAS CLEARLY IDENTIFIED DIFFERENT REGIONS AND COLORS IN A 
DIFFUSION FLAME. 



Highly luminous zone (bright 
yellow); continuous spectrum 
due to thermal emission of hot 
carbon particles on the fuel 
side; visible radiation Is 
Increased due to soot. 

HYDROCARBON FUEL DECOMPOSES 
TO SIMPLER PRODUCTS 
(ESPECIALLY HYDROGEN) AND 
DIFFUSES TOWARD OXYGEN REGION. 

The bright yellow region Is 
edged at the base of the flame 
by this bluish green region 
showing C2 and CH emissions; C2 
bands aDDear green (shaded toward 
violet) and CH bands aDDear violet 
blue to the eye. 



i 



Combustion products (H 2 0,CO,C02) 
exist In the entire region. 




Darker central region; OH 
emission and absorption In 
the UV and vlb.-rot. bands 
In the IR, but little 
visible emission. 

Luminous zone (blue flame); 
soectrum of burning CO and 
sometimes bands of the 
Schumann-Runge system of O2 



DIFFUSION 

outer region show- 
the soectrum of burning CO 



t 



1 

I 

/WALL 

\ 

I 

: 



H 



BURNER 



— ScmK»Af]plicatia u titmimtion^Corfxy a ) K i)- 



RATIONALE FOR OBSERVING THE COLOR OF FLAMES (CONTB) 

I FLAMES EMIT DISCRETE BAND SPECTRA. 

• SPECTRA IN VISIBLE AND UV REGIONS ARE DUE TO CHANGES OF ELECTRONIC ENERGY IN 
NEAR IR ARE DUE TO CHANGES OF VIBRATIONAL AND ROTATIONAL ENERGY, AND IN EXTREME 
IR ARE DUE TO CHANGES IN ROTATIONAL ENERGY ONLY. 

• NONE OF THE STABLE PRODUCTS OF COMBUSTION (I.E., H,0, CO,, CO, 0, OR N„> GIVE 
SPECTRA OF APPRECIABLE STRENGTH IN THE VISIBLE OR UV REGIONS. 

• ONLY OH GIVES A BAND SYSTEM IN THE UV. 

• FOR HOT FLAMES, THE BULK OF THE ENERGY IS RADIATED IN THE CO, BANDS AT 2 7 
AND 4. ^ AND FOR WATER VAPOR AT 2.8 M . 

• OTHER VIBRATIONAL BANDS OF INTEREST ARE CO AT 2.3 AND U.5 M , NO AT 2 6 AND 
5.2 M - AND THE OH VIBRATIONAL BANDS THAT COVER THE NEAR IR AREA TO ABOUT Hu . 

• AT FLAME TEMPERATURES, THE MAXIMUM EMISSION TENDS TO BE IN THE NEAR IR 
BETWEEN ABOUT 2 M AND 1 M ACCORDING TO THE TEMPERATURE OF THE FLAME. 

I FOR CLEAR FLAMES, THE RADIATION IN THE VISIBLE AND UV ACCOUNTS FOR LESS 
THAN 0.« OF THE HEAT OF COMBUSTION. THIS VISIBLE RADIATION COMES MAINLY 
FROM THE INNER CONE, WHILE THE IR RADIATION COMES FROM THE MAIN BODY OF THE 
GASES, BOTH THE INTERCONAL GASES AND THE BURNT PRODUCTS. 

• USING BLUE, GREEN OR YELLOW LIGHT, THE PHOTOGRAPHS OF BUNSEN FLAMES NAVE 
SHOWN ONLY THE INNER CONE CLEARLY, BUT USING INFRARED LIGHT OF WAVELENGTH 
8,500 OR 10,050 °A, THE INNER CONE IS INVISIBLE BUT THE INTERCONAL GASES 
SHOW UP. 

• FOR LUMINOUS FLAMES, RADIATION FROM SOOT PARTICLES INCREASES THE RADIATION 
IN THE VISIBLE. 



Science Applicslions International Corporation - 



JUSTIFICATION FOR A SPACE EXPERIMENT 



• < 10~ 3 g is required 

• g-jitter at f < 1 Hz is required 

• Test times » 40 sec 



. Sconce A&katkxvHtmatmal Corporation- 



ROLE OF ADDITIONAL GROUND-BASED EXPERIMENTS 

• OPTIMIZE THE FINAL SPACE EXPERIMENT TEST MATRIX 

I MINIMIZE THE SPACE EXPERIMENT COMPLEXITY BY REDUCING THE MEASUREMENT 
REQUIREMENTS 

• ASSIST IN THE DEVELOPMENT OF TECHNOLOGY FOR THE SPACE EXPERIMENT 



i Science Applications Alternations Corporation - 



OVERALL PROGRAM APPROACH 



■ Science Applications titamational Corpora ti on - 



OVERALL PROGRAM 



CODR 

-« — 



PRR 



CODR — PRR (2 YEARS) 



PRELiniNMT S-SEC 

0-« TESTS! 
I ESTABLISH NUMB 

OFFERED IT WOP TOMER 
I OBTAIN PHOTOGRAPHIC DATA 
I DEFINE K-135 TESTS 




ADDITIONAL S-SEC 
0-« TESTS (VI TH 
DIAGNOSTICS) 

I DATA FOR SCIENCE 

RE0U1 REPENTS 
I OPTIMIZE KC-13S ADD 

SPACE TESTS 
< TESTING OF DIAGNOSTICS 



PRELIMINARY tt-135 
0-« TESTS! 
• ESTABLISH BOUNDS 
OFFERED IT tt-135 
t OBTAIN PHOTOGRAPHIC 
DATA 





ADDITIONAL K-135 
0-« TESTS (KITH 
DIAGNOSTICS) 

t OBTAIN DATA 

* DEFINE TESTS FOR SPACE 




REPORTt 
• DROP-TWI RESULTS 
I KC-135 RESULTS 
I ANALYSES 
I DEFINITION OF 
SPACE EVER I RENT 



SPACE 
EXPERIMENT 



r 



Saa&ApplkaOomHvnationalCorpontion ■ 

ACTIVITIES AFTER COVR AND BEFORE PRR 



• 



• 



i 



INITIATF THE DETAILED PLAN FOR EXPERIMENTS IN THE 5-SEC. DROP-TOWER FACILITY, 
f I J TH EXPE RJMENTAL C AMBER AND 1GNIT0R, AND CONDUCT DROP-TOWER EXPERIMENT 
TODE^LOP THE X VERIFY IGNITION. AND OBTAIN INITIALLY ONLY PHOTOGRAPHIC 
AND THEN COMPLETE DATA. 

INITIATE THE DETAILED PLAN FOR KC-135 TESTING. DESIGN ^ EXPERIMENTAL CHAMBER 

[m CONDUCT KC-135 AIRCRAFT FLIGHT TESTING FOR NOMINAL O-G TEST TIMES OF 

l! 15 X Hs5s TK PERFORMANCE OF THE EXPERIMENTAL HARDWARE AND ANALYZE 

THE TEST DATA RESULTS. 

FINALIZE THE SPACE EXPERIMENT TEST MATRIX BASED ON THE RESULTS OF GROUND-BASED 

EXPERIMENTS. 

INITIATE THE NUMERICAL ANALYSIS FOR TRANSIENT IGNITION AND STEADY-STATE 
Si UNDER MICRO-GRAVITY CONDITIONS *»™ ■^J,^^'^ 
CONDITIONS UNDER WHICH THE EXPERIMENTS A E GOING TO BE CO DUCTEI BY USING 
ELLIPITIC EQUATIONS IN CONJUNCTION WITH FINITE-RATE KINETICS. SOOT FORMATION 
AND RADIATION. 



mm J 



Science Applications ktomuknel Corporation - 



GROUND-BASED EXPERIMENTS 



I Science Applications htemational Corpora* k 



GROUND-BASED EXPERIMENT COMPONENTS 



EXPERIMENTAL CHAMBER 

CAMERA 

IGNITION SYSTEM 

THERMOCOUPLES 

PRESSURE TRANSDUCER 

FUEL TANK 

FUEL LINES 

VALVES AND FITTINGS 

RADIOMETER 

SAMPLING PROBES 

SAMPLING BOTTLES 



— ScmrKeApplk^oorakVamadonal Corporation- 



RESULTS OF GROUND-BASED TESTING 



• Improve theoretical model 

• Develop experimental technique 

• Optimize space experiment 



, Science Applications International Cunjotation ■ 

r EAR THSOUND ANV SPACE PROGRAMS 

-JITTER AT f<1Hz; TEST TIMES = 40 sec. 



REQUIREMENT.^ 



EXPERIMENTS 
WILL STUDY: 



MEASUREMENT 
TECHNIQUES TO 
BE USED: 



<10" 3 GJ 



EARTHBOUND PROGRAM 



DROP TOWER: 

T=5 SEC. 
G=0(10" 5 )- 

G-J1TTER NEGL. 



KC-135: 
t=10-15 sec, 
G~10~ 2 IF FREE FLOATING/ 
CABLE MAY CAUSE JITTER. 



SPACE PROGRAM 



• IGNITION 

• TRANSIENT FLAME DEVELOPMENT 

• STEADY-STATE FLAMES 



SHUTTLE: 
t > 40 sec. 

g= <10" 3 AVAILABLE. 
TEST DURING 'QUIET 

TIMES*. 



• PHOTOGRAPHY 

• TEMPERATURE MEASUREMENTS 

• RADIOMETRY 

• COMBUSTION PRODUCTS SAMPLING 



• IGNITION 

• TRANSIENT FLAME 
DEVELOPMENT 

• STEADY-STATE 
FLAMES 



PHOTOGRAPHY 

TEMPERATURE 

MEASUREMENTS 

RADIOMETRY* 

COMBUSTION 

PRODUCTS 

SAMPLING 



'IF EARTHBOUND EXPERIMENTS SHOW THAT RAD IATION MEASUREMENTS ARE ''"f'Sr. ^cSISts 
OF ThSeTcaI ANALYSES AND EARTHBOUND RESULTS WILL BE USED TO DELETE THOSE MEASUREMENTS 

FROM THE SPACE PROGRAM. _____ JSTOT/^"- 

___ ■ ' -__&*■■ -_#- 



Sdarca Applications kuarnathnal Corporation- 



CONCEPT DESIGN 



. — Soan» Applications hlarnational Corporation- 



PROCEDURE FOR PREUMIHARY STUDIES 
M THE 5-SEC DROP TOWER 



• START THE CAMERA AND DROP THE TEST CELL 

• START THE FUEL FLOW AND IGNITION SYSTEM FROM 1/2 
TO 1 SEC AFTER RELEASE OF THE TEST CELL 

• DURATION OF SPARK IGNITION TO BE OPTIMIZED WITH TESTING 

• SEQUENCE OF OPERATIONS CONTROLLED BY BATTERY-POWERED TIMER SWITCH 

I MOVIE CAMERA RECORDS BOTH THE I6NITION PROCESS AND FLAME DEVELOPMENT 

I THESE RESULTS WILL BE USED IN FINALIZING THE KC-135 FLIGHT TESTS 



sd 



I Saanca Applications titemational Corporation- 



PROGRAM SUMMARY 



c-£ 



WtPARISOHflimEEN MIP-PECK, SPACELAl IMP MATERIALS 
SCIENCE IAS FOR CONPUCTING THE EXPERIMENT 



I MID-DECK 
• SPACELAB 



VOLUME AND WEIGHT CONSTRAINTS (-) 
DIRECT ASTRONAUT INVOLVEMENT (+) 
LESS COMPLEXITY (+) 
LESS AUTOMATION (+) 
VENTING (-) 

MORE AVAILABLE VOLUME (♦) 
DIRECT ASTRONAUT INVOLVEMENT <♦> 
LESS COMPLEXITY (♦) 
LESS AUTOMATION (♦) 
VENTING (♦) 



MATERIALS SCIENCE LAB 






MORE AVAILABLE VOLUME <♦) 

INDIRECT ASTRONAUT INVOLVEMENT (-) 

MORE COMPLEXITY DUE TO FULL AUTOMATION (-) 

EXPOSURE OR LACK OF EXPOSURE TO SUN MAY CAUSE PROBLEMS IN RELATION 

TO COOLING OR HEATING OF THE CHAMBER (-) 

VENTING (♦) 



SammAppHcatiocstntetnational Curpoffon- 



COMPARISONS BETWEEN MID-DECK, SPACELAB AND 
MATERIALS SCIENCE LAB TOR CONDUCTING THE EXPERIMENT 





MSL 


MID-DECK 


SPACELAB 


VOLUME CONSTRAINT 


♦1 





♦1 


WEIGHT CONSTRAINT 


+1 





+1 


DIRECT ASTRONAUT INVOLVEMENT 


-1 


♦1 


♦1 


COMPLEXITY DUE TO AUTOMATION 





+1 


♦1 


COMPLEXITY DUE TO DESIGN 


-1 


♦1 


♦1 


VENTING 


+1 


-1 


♦1 


2: 


♦1 


♦2 


♦6 



I Science Applications hlemation^Coipcratk. 



EXPERIMENT DESCRIPTION 



• COMPONENTS : 

NOZZLES. EXPERIMENTAL CHAMBER. FUEL TANK. FUEL VALVES. 
THERMOCOUPLES. TRANSDUCER. SAMPLING PROBES. SAMPLING 
BOTTLES, IGNITOR. HEAT MANAGEMENT SYSTEM. CAMERA(S). 
DATA ACQUISITION SYSTEM. 



I FUEL: 

METHANE; PROPANE 



-Science Applications International Corporation- 



EXPERIMENTAL CHAMBER 

SEALED CHAMBER. EITHER CYLINDRICAL OR RECTANGULAR. ALUMINUM 

MUST BE ABLE TO WITHSTAND THE SPECIFIED CONDITIONS FOR THE REQUIRED PERIOD 

OF TIME FROM COMPLETION OF ASSEMBLY TO POST-FLIGHT DATA ANALYSIS 

MINIMUM HEIGHT/DIAMETER (OR WIDTH) - 1.5 

VOLUME (> 0.0*1 m 3 ) 

TWO WINDOWS (ONE FOR CAMERA. ONE FOR OBSERVATION) 

PORKS) FOR THERMOCOUPLES. SAMPLING PROBES. IGNITORS. AND TRANSDUCER 

NUMBER OF NOZZLES ENTERING THE CHAMBER: TWO 

CHAMBER MUST WITHSTAND PRESSURES OF UP TO 3 ATM. GAS TEMPERATURES OF UP 

TO 495K. AND METAL TEMPERATURES OF UP TO 310K (CONSERVATIVE ESTIMATES FOR 

UPPER LIMITS) 



r 



Science Applications International Corporation - 



VIEWING 



• THE FIELD OF VIEW MUST BE 5 - CM HORIZONTAL BY 10 - CM VERTICAL AT 
THE PLANE PASSING THROUGH THE FLAME CENTERLINE AND PERPENDICULAR TO 
THE AXIS OF THE CAMERA 

| BESIDES VIEWING PORT FOR CAMERA. A VERTICAL VIEWING SECTION MUST BE 
PROVIDED FOR OBSERVATION (EITHER DIRECT OR VIA A TV CAMERA) 



SdencmApplicatkjmkitmmationalCorporatiori- 



WINDOWS 



TOTAL NO. OF WINDOWS 



VIEWING PORT(S) 



OBSERVATION WINDOW 



EXPERIMENT IN SPACELAB 



1; USED FOR HIGH-SPEED CAMERA; 
SMALLER THAN OBSERVATION WINDOW. 



1; LARGER THAN (AND 90° APART FROM) 
THE VIEWING PORT FOR HIGH-SPEED 
CAMERA; USED FOR ASTRONAUT'S OBSERVATIO 



— Science Applications International Corpontk 



GAS COMPOSITION IN THE EXPERIMENTAL CHAMBER 



STANDARD AIR (211 2 . 79Z ty. ALSO. 10Z 2 -90Z H ? . 

OXYGEN DEPLETION: <10Z AT THE END OF THE EXPERIMENT. 

USE PREFILLED CHAMBER MIXTURE. 

2 TOLERANCE AT THE BEGINNING OF THE TEST SET: ±0.1 MOLE Z. 

INITIAL PRESSURES FOR EACH FLAME - 1 ATM AND 1/2 ATM. 

GAS TEMPERATURE BEFORE EACH TEST - 295 ± 5K. 

PRESSURE TOLERANCE - ±0.3 PSIA BEFORE EACH TEST. 

CHAMBER PRESSURE SHOULD BE RECORDED BEFORE. DURING AND AFTER THE 

EXPERIMENT (AT LEAST EVERY 50 MILLISEC. ..... 20 SAMPLES/SEC ) 

CHAMBER TEMPERATURE MUST BE MONITORED DURING THE ENTIRE EXPERIMENT 
(AT LEAST EVERY 50 BILLISEC. ,. e .. 20 SAMPLES/SEC.) 



r 



Science Applications International Corporm 



mim 



ition- 



VENTING 



ADVANTAGES 



t 



MANY FLAMES CAN BE STUDIED IN ONE FLIGHT 

SMALLER CHAMBER 

WEIGHT REDUCTION 

CONTROL OVER THE Z 2 BURNED 

FLAMES WITH DIFFERENT CHAMBER PRESSURES CAN BE STUDIED 

NO CHAMBER COOLING AFTER EACH FLAME 

IF FUEL IS ACCIDENTALLY RELEASED IN THE CHAMBER. EVACUATION IS EASY 



DISADVANTAGES 

• AIR MUST BE CARRIED ABOARD 

• VALVES FOR EVACUATING AND RE-FILLING THE CHAMBER 

• ACCURATE CONTROL OVER CHAMBER PRESSURE 
I WEIGHT ADDITION DUE TO AIR TANK 



Saence Applications International Corporation - 



FUU-WBIHG SySTEM 



• STAINLESS STEEL TUBES 

• FLOW CONTROL ACCURACY = ♦ 0.01 CM 3 /SEC 

I FLOW VALVES, EXPLOSION - PROOF SOLENOID VALVES. 
AND REGULATORS 

• PRE-CALIBRATED PRECISION VALVES 






Saence Applications International Corporation - 



NOZZLES 



• RADIUS = 0.05 - 0.15 CM 

I DISTANCE FROM PORT TO CHAMBER CEILING 2:30 CM 

• DISTANCE BETWEEN NOZZLES (IF MORE THAN ONE NOZZLE PER CHAMBER) - 5-10 CM 

• NOZZLE(S) EQUALLY DISTANT FROM CHAMBER CENTERLINE 
I NOZZLE(S) IN LINE WITH THE CAMERA 

I MINIMUM DISTANCE FROM WALL FOR NOZZLE POSITIONING - 10 CM 

• THIS NOZZLE ARRANGEMENT ENSURES THAT TEMPERATURES S350K AND NEGLIGIBLE 
GAS VELOCITIES EXIST AT THE CHAMBER CEILING (i.e.. PREVENTION OF EXTENSIVE 
HEATING Of CHAMBER CEILING AND SIGNIFICANT RECIRCULATION) 

• PROXIMITY OF THE NOZZLES DOES NOT CAUSE FLAME DISTURBANCE BECAUSE MAXIMUM 
FLAME RADII =0.5 - 1.0 CM 

• NOZZLES MUST BE SECURED INSIDE THE CHAMBER TO AVOID MISALIGNMENT OR LOOSENING 
DUE TO VIBRATIONS OR DURING LIFT-OFF 

• MINIMUM NOZZLE LENGTH IN THE CHAMBER - 15 CM 



I Science Applications htmaiional Corporation - 



mi 



PURE. HIGH-OUALITY METHANE 

PROPANE WILL BE CONSIDERED FOR LATER FLIGHTS 

METHANE VOLUME FLOW RATE ■ - 15 CM 3 /SEC 

PROPANE VOLUME FLOW RATE - - 5 CH 3 /SEC 

AMOUNT OF FUEL CARRIED WILL BE SUFFICIENT TO PERFORM THE 

ENTIRE TEST MATRIX 

FUEL TANK VOLUME - TBD BASED ON 

• NUMBER OF FLAMES 

• VOLUME FLOW RATE OF EACH FLAME 

• 40-SEC BURNING TIME PER FLAME 

• PRESSURE IN THE FUEL TANK 



— i>c>encm Applications titemationait^crpcrMion- 



mt TANKfSJ 



INITIAL PRESSURE » 5 - 15 ATM 

NO FUEL(S) LEFT AT THE END OF THE EXPERIMENT 

STAINLESS STEEL TANKS 

DUAL SYSTEM FOR PRESSURE CONTROL 

FUEL PRESSURE REMAINS FAIRLY CONSTANT THROUGHOUT THE EXPERIMENT 

ONE FUEL TANK IF ONLY CH,,: TWO FUEL TANKS IF BOTH CH„ AND CjHg 




P 




Bgroyg rxreniMgwr 






AFTC* EXFgRIMEMr 



-/ 



ScienceApplicationslntemationalCcrparation- 



IGNITION 



MINIMUM DISTURBANCE TO FLAME 

EITHER SPARK OR HEATED-WIRE IGNITION (SPARK PREFERRED) 

POWER IS SUPPLIED BY SHUTTLE POWER 

IGNITION AND FUEL INJECTION MUST BE SYNCHRONIZED 

FINAL DETAILS OF THE IGNITION SUBSYSTEM TBD FROM THE 

5- SEC DROP-TOWER TESTS 






Science Applications International Corporation - 



IGNITION SYSTEM OPTIONS 



SPARK IGNITION 1 



• Nozzle is treated 
as an electrode. 

• Electrodes are fixed. 

• Power Is supplied 
by shuttle 



fi 



HEATED-WIRE IGNITION 



•Provisions must be made for 
appropriate coils to increase 
the spark voltage. 

• Interference with other Shuttle 
operations must be addressed. 



• One ignitor is sufficient for both nozzles if 
the ignitor arm Is 
rotatable. 

• Ignitor can have only 
two positions (Le., 
vicinity of the two 
nozzles). 

• Once Ignition occurs, 

the arm is rotated and the ignitor is moved 
to the other nozzle location. 

• Ignitor Is coiled 0.33-cm-diameter nlchrome 
wire. 

• Power Is supplied by shuttle 




• Ignition and fuel Injection must be synchronized. 

• Final details of Ignition subsystem TBD from the drop tower tests. 



* Preferred 



I Scone* Applications titamalronal Corporation - 



HEAT RELEASE PER FLAME 



CH^ + ZOz—- CO^f 2H 2 



C 3 H g + 5-qj-*3C0 -t-4-HjO 



w *%-*(-?¥■)- te&^) 



WHERE o - ENTHALPY OF COMBUSTION OF FUEL AT 25°C WITH WATER VAPOR IN THE PRODUCTS, 
«F - MASS OF FUEL, t = EXPERIMENT TIME, V p = VOLUME FLOW RATE OF FUEL. T Q • INITIAL 
TEMPERATURE. P Q « INITIAL PRESSURE. Mp « FUEL MOLECULAR WEIGHT. AND fi - UNIVERSAL 
GAS CONSTANT. 

FOR CHj,: o » 50.010 kJ/kg. M = 16 kg/kgmole 
FOR C 3 H 8 : o » 16.553 kJ/kg. N - 44 kg/kgmole 

USING T - 300K. R = 8.3143 kJ/kghole-K AND t - 40 sec. WE OBTAIN THE FOLLOWING 
DIAGRAM: 



- Sonne* Applications kHemational Corporation - 



HEAT RELEASE PER FLAME (CONT.J 

(Q releose^ - 0.Q128 PV p (0 ^^^ - 0.0326 PV F with 
Q reieose In kJ, P In kPa and Vp In an 5 /sec. 



18 










16 
1 










1 M 




eft / 


c 3"8 y 




° 12 




(1 amy 


kWQM/r 


y^ ®t> 


a in 








y (1 atm) ^ 


Heat Rele 

en ca 








-^- ^^ ^ 


4 








(1/2 otm) 


2 




i i i 


i i i 


Ewrlment time - 40 sec 

J ]_.. 1 1 1 1 1 l 



6 8 10 

Fuel Flew Rate. V F (cm 3 /sec)— 



12 



14 



16 



SdencaApplicatianslnteiTiationalCcrparatian- 



CHAMBER TEMPERATURE 



TEST CELL TEMPERATURE SET POINT FOR EACH FLAME - 295 

i 5K 

TEMPERATURE RISE AT THE END OF EACH FLAME MEASUREMENT 

- TBD (TYPICAL 20 - 200K MAXIMUM) * 

COOLING BY FAN (IN SPACELAB) 



GAS TEMPERATURE WITH ADIABATIC WALLS 



, — SciencaApplkMions International Coloration- 



TEMPERATURE RISE IN THE EXPERIMENTAL CHAMBER 



ASSUMPTIONS ; Ideal gas 

All of the heat release Is absorbed by the gas only 

No heat loss 

Constant specific heat 

Burning has negligible effect on no. of moles of air 



Fran the 1st low of thermooVnanlcs, 

where we have used the previously-derived expression for V^, (see Appendix). 
Eauatlng this relation with the previously- derived expression (see Appendix) 



yields 



AT ■ T-7T - ^Mpt/c^+piH^ 




In this derivation- u - specific internal energy and Cy - constant-volum specific heat. 



I — Science Applications International Corporation - 



TEMPERATURE RISE IN THE EXPERIMENTAL CHAMBER (CONTJ 

Using ( Cv ) Qlr = 0.718 Id/kg - K, (c^ « M « =0.732 kJAg - K, 

f\j ir = 28.84 kgAg role, ^ „ 10I n = 28.40 kg/kg role, 

aid also 

for CH q : Mp = 16 kgAg role, 1 = 2, Op =• 50,010 kJ/kg, 

for CjHg: r\c = W kgAg role, 1 = 5, qp = 46,353 kJAg, 
we obtain the gas-tenperature rise as a finctlan of I oxygen burned, 

1000 



800 



600 



400 



200 



' P » 1/2 otjn or 1 atm 




■ Exn. time » 40 sec 




C^/alr 




: \ 


C 3 Hg/90X Nj-lOJ O2 ^^^~>~ 


\ 


^^ggsss^^'^^^ —* 


^•e^^^^^^^ ssS=S: 


CHj/9(K Hj-iaZ O2 — / 



0.02 0.04 0.06 0.08 

x=n 02, burned/"^, Initial"* 



0.10 



I — Science Applications International Corporat i on - 



CHAMBER PRESSURE 



• TEST CELL PRESSURE SET POINT FOR EACH FLAME - 14.7 ± 0.3 
PSIA - 101 t 2 KPA - 1 ± 0.02 ATM; ALSO, 0.5 ♦ 0.01 ATH FOR 
SOME FLAMES 

• FOR ALL TESTS. TOLERANCE ON PRESSURE SET POINT IS ♦ 2X. WHETHER 
P - 0.5 ATH OR 1.0 ATM 

• PRESSURE RISE DURING COMBUSTION: 

• STRONGLY DEPENDENT ON CHAMBER VOLUME 
I TBD 

• TYPICAL: 51 (WIN.) TO 80Z (MAX.) 

• TRANSDUCER MEASUREMENT PRECISION - ± 0.15 PSIA (i.e., IX) 



, Science Applications 



Wemrtiona/Corpor»fion 



"pressure rise in the experimental chamber 



t v = 71 n r 



Z. - X 



^P- 2L AT 



USING THE RELATION DERIVED FOR AT. 



2.0 
1.5 



E 

s l.o r" 

O- 

<1 



p - 1/2 atnt Exp. tire * 40 sec 

A: CjHg/alr 

B: C 3 Hg/90X Nj - 105 Oj 

C: CH 4 /air 

D: CH^/901 Nj - 101 O2 



0.02 0-04 O- 06 

X"^ burned/ 1 ^' initial" 



0.08 



, — Science Applications 



kiamalional Corporation - 




0.10 



PRESSURE RISE /If THE EXPERIMEHTAL CHAMBER (CONT.) 



P - 1 mm; Exp. tire « MO sec 

3.0 [ A:( W alr 
t L B: CjHg/SB Nj - IK 0j 

1 20 [ C:CH * / ° lr 

I ' [ D: QV90X N2 " llK °2 



0.06 



0.02 0.04 

X " \, tuned/ 'V Initial" 



0.06 




— SduncmA&kMioni International Caponthn- 



COOLING Of THE CHAMBER 



ZnwlZV "'"' A " E " f ™'" E ° F ° R ™ E CN «" Em '"™™ ™ c* 
orspaIab, T " "*' " C " RMt ""**» " F '» «*« 

THE*,,™ GAS-TEHPERATK R,SE IK THE CHAWR » IU tt , 100 K ra „ ^ 

*a o, K «t,o» » c jV „» a«. «« „ „ EAT nmm T0 rHE Alm , M 

^ S , f T,T " S ;™ ,r » RE «' SE » 1LL «« « MIMB TE*ERAT« BY 
<K IF AIL Of THE HEAT RELEASE IS TRANSFERRED TO ItTAL 

2:\™:z ,un ™ !Y pmv,di " g ,K "^ ** ™ ««• «■*«« 

SPECIFICS OF COOLING: TBD. 



_ 



SoenceApplkatkxvhtematmalCcrporwion- 



WATER - VAPOR CONDENSATION 



• We are solving for on amount of fuel that when 

* r ; ed ' nt ° C0 2 "d ^ results in condensation 
(relative humidity. * = p w / D = 1} at a cnamber 

Pressure of 1 atm and the'lowest allowable chamber 
temperature of 290K. "wver 

• CH„ ♦ 20 2 - C0 2 ♦ 2H 2 ; CjHg ♦ 50 2 - 3C0 2 ♦ ^0 . 

• At the end of the experiment, assuming that m CH and m r u 
.rans of fuels are burned, the total nur*>er of Soles ft chanter is 

n t*f " V + ( V ^V + In C0o + ln M 



n N„l 



+ (n 



'o 2 ,i 



- 2n 



2-- "2'" ~ <-"2 ~ n 2 

„ " 5n C 3 H 8 ' + <" CHj< ♦ 
" n alM *"ch +2 n CH 




'CH, 



'« - 'Vs' + % + '"C 3 H ) * <2n C u ♦ Hac„, 



"3 n 8 



where subscripts t - totolA - nnoL i - initial, and b - burned. 
But 



P H,0 



n. 



'H,0 



n t,f 



2n CH„ * "n, 



'r h 
C 3 H 8 



"air. I +n CH (l + 2n CH , 



n alrjl /P.(2/P H20 -i/P, ( n CHjj + 2n ^ Hfl) 



'« " L 3"8 
•3"8 } 



Scenes Application International Corporation 

WATER 'VAPOR CONDENSATION (CONTINUED) 

• Therefore A IU . 0m /M x 

V ch , = « T (2/p H20 - 1/P) (^CH/H + 2m C 3 H 8 /M C 3 H 8 ) 

. using T - 290K, P - 101.325 kPo, P „ * P g = 2.0855 kPa (fro. soturated H 2 tables 
f0 r T - 290 K,. *„, - 16. » C ,H R = «* "'K^' «» * = 8 - 31W kJ/kfl "° ,e - K ' * 



obtain 

v ch. - W m CH 4 + m m C 3 H 8 



( m in kg, V In m 3 



chanter for a chamber environment of P - 1 atn. and T 290 K. 



ScierK»Appl>catiomMarnationalCorporation 



MEASUREMENTS 

• Flame Shape Development and Rame Extinction 

• Color and Luminosity 

• Temperature Profiles 

• Species Concentrations 

• Radiation Measurements 
•Chamber Pressure 
•Acceleration Environment 



I Science Applications international Corporation 



FLAME-DIMENSION MEASUREMENT 



• A SCALE FACTOR FOR THE FLAME-LENGTH MEASUREMENTS CAN BE OBTAINED 
FROM THE BURNER DIAMETER. 



d 



— Scienca Applications International Corporation - 



• 



PHOTOGRAPHY 



• ONE CAMERA PREFERRED' 

• COLOR, 16nm FILM 
MUST HOLD ENOUGH FILM FOR UP TO 10 SEC 

• FRAMING RATE - 400 FRAMES/SEC FOR THE FIRST 5 SEC, OF MEASURE- 
MENTS (TO STUDY IGNITION) AND 100 FRAMES/SEC FOR THE REST OF 
THE PERIOD 

• A TWO-SPEED CAMERA IS PREFERRED 
CAMERA STARTS BEFORE FUEL STARTUP AND IGNITION 

• LENS-TO-NOZZLE DISTANCE - TBD 

• IF TWO NOZZLES ARE USED, LENS-TO-NOZZLE DISTANCE MUST BE 
ADJUSTED 

• FILM CHANGE AFTER EACH FUME MEASUREMENT 

• TIMING SYSTEM SHOULD BE PROVIDED TO RECORD ACCURATE TIME 
INTERVALS (-0.1 SEC) ON THE FILM 



• 



TWO CAMERAS IF LIMITATION ON FILM LENGTH OR DIFFICULTIES WITH TWO-SPEED CAMERA. 



. Science Applications 



htemational Corpora*** - 



THERMOCOUPLES 



, RAKE OF THERMOCOUPLES FOR MEASUREMENTS OF TEMPERATURE DISTRIBUTIONS; 
MINIMUM: 6 THERMOCOUPLES. PREFERRED: 9 THERMOCOUPLES PER NOZZLE. 

t GAS TEMPERATURE MEASUREMENTS CONDUCTED AT FIXED LOCATIONS 
ABOVE THE FLAME (5-15 cm ABOVE THE NOZZLES). 

■ EXAMPLE- TYPE 'V Pt/Pt-10X Rh (1650°C UPPER LIMIT. 1760°C 

MELTING POINT). WIRE DIAMETER - 0.13 m. BEAD SIZE - 0.« m, WK > CAN 
BE EXPOSED. OUTPUT VOLTAGES (16 t 0.1 *> SHOULD PROVIDE TEMPERATURE 
MEASUREMENTS WITH i 10K. 

• MEASURED ACCURACY VILL NOT POSE A MAJOR PROBLEM SINCE CORRECTIONS 
FOR HEAT CONDUCTION. RADIATION. ETC. ARE AVAILABLE. 



SAMPLING RATE - 20/sec. 



SSWL: 



- Science Applications International Cutporatkm 

THERMOCOUPLES (CASE A) 



viewing 
Port 2 



L — carerol 




Vlewlre Port 1 



Viewing Port 
for TV arera 
or Astronaut 



DIMENSION NOT TO SCALE; 
SEPARATION DISTANCE x - TBD; 
ONE CAMERA IS PREFERRED. 




Science Applications htematianal Corporal 



if ion- 



THERMOCOUPLES (CASE ff/ 



Viewing 
Port 2 



Canera 2 



Viewing Port 1 




Viewing Port 
for TV canera 
or Astronaut 

DlfENSION NOT TO SCALE; 
SEPARATION DISTANCE x * TBD; 
ONE CNERA is PREFERRED. 



* Preferred 



Sdancm Applications ttfamaf iansJ Corponttk 



GAS SAMPLING 



QUENCHING QUARTZ NICROPROBES 

5-10 c* ABOVE THE BURNER PORT (-0-5 c ABOVE THE FLAW TIP) 

^ZlS^ DURAT, ° N) ^ * SPEC ' F,ED TlflE INT ° mPUm UHKS (VA ™ 

THREE PROBES FOR EACH BURNER (RAKE OF PROBES) 

ONE TANK PER PROBE PER FLAME (VOLUME OF TANK -30 ch 3 ) 

SAMPLING BOTTLE INITIALLY EVACUATED 

FINAL PRESSURE IN THE SAMPLING TANK: -5 - 7 PSIA 

DETAILS TBD 

SAMPLING CONDUCTED WHEN FLAME REACHES STEADY STATE 

PERTURBATIONS DUE TO SAMPLE REMOVAL DO NOT CAUSE FLA* DISTURBANCES AS LONG 

AS SUCTION IS NOT SUDDEN 



Soencs Applications International Corporator! - 



RADIOMETER (COUT.) 



o 
i 
• 



RECTANGULAR APERTURE = 11.5 cm HEIGHT x 1 cm WIDTH 

BACKGROUND IS PROVIDED BY THE LIGHT TRAP (HEDGE ANGLE = 12°) 

THE INTERNAL APERTURES AND ADJUSTABLE EXTERNAL RECTANGULAR APERTURE PREVENT 

ERRONEOUS READINGS DUE TO INTERNAL REFLECTIONS OR UNWANTED STRAY RADIATION. 

THE SHUTTER PROVIDES ZERO-OFFSET READINGS THAT CAN BE SUBTRACTED FROM THE 

OPEN-SHUTTER OUTPUTS. 

THE TOTAL RADIANT POWER Q RAD CAN BE COMPUTED FROM THE MEASURED IRRADIANCE H 

UNDER THE ASSUMPTION OF SPHERICALLY-ISOTROPIC EMISSION AS Q RAD - 17TS 2 H 

WHERE S IS THE DISTANCE BETWEEN THE RADIOMETER SENSOR AND FLAME AXIS. 

DISTANCE FROM NOZZLE CENTERLIKE TO THE SENSOR - 30 cm. 

THERMOPILE SENSOR: E.G. SENSORS INC. MODEL CI WITH INTRA* 2 WINDOW. SENSITIVE 

AREA 1 mm DIA. 

A PROGRAMMABLE CALCULATOR CAN BE USED FOR DIGITAL RECORDING OF THE 

ELECTRICALLY-FILTERED RADIANCE OUTPUT. 

DETAILS TBD BASED ON FURTHER STUDIES. ANALYSES AND ENGINEERING CONSTRAINTS. 



♦BASED ON THE DESIGN OF MARKSTEIN: ™TH SYMPOSIUM (INT'I ) ON COMBUSTION. PP. 10551061. 



Science ApplicatinnslntemationalConjoration- 



POWER REQUIRED FOR EACH FLAME MEASUREMENT 

(CONSERVATIVE ESTIMATES) 



SPACELAB AVAILABLE POWER: > POWER IN MSL. 



A 
B 

C 
D 
E 
F 
G 
H 
I 



COMPONENT 



16nm CAMERA 

IGNITCR: NICHROME WIRE (OR 
SPARK) 

SOLENOID VALVE FOR FUEL 

EXPERIMENT CONTROLLER 

6-9 DIGITIZED THERMOCOUPLES 

FAN (CHAMBER) 

CONTROL PANEL 

TEMP. & PRESSURE CONTROLLER 

3 VALVES FOR SAMPLE REMOVAL 



POWER (WATTS) 



60 

50 (OR NEGL.) 

10 

10 

20 

30 

10 

10 

15 



TIME (SEC) 



HO SEC PER FLAME 

2-5 SEC TBD (OR NEGL.) 
HO SEC 

ALL THE TIME 
> MO SEC 
TBD 

ALL THE TIME 
ALL THE TIME 
<5 SECj TBD 



SOURCE 



165-215 



ORBITER 

ORBITER 
ORBITER 
ORBITER 
ORBITER 
ORBITER 
ORBITER 
ORBITER 
ORBITER 



•RADIOMETER AND ACCELEROMETER POWER REQUIREMENTS NOT INCLUDED. 



^J 



- Science Applications International Corporation - 



GAS -SAMPUNG CONFIGURATION 



flUWTZ 
TKOUS 



cumr i W " L 




Separation distance (!): TBD 



- Science Applications International Contortion - 



nmet wrnis 
Tw*. ruini. t«* 

K«TLK 



RADIOMETER* 






Experimental Chamber 



Wide-angle radiometer 
Shutter 



window for Camera 



Wedge-Shaped 
Light Trap 




Thermopile Sensor 
Circular Apertures 



Rectangular Aperture 

Radiometer Support 
i — Camera* 



Window for Camera 



Nozzles 
Observation Window 



• BASED ON THE DESIGN OF MARKSTEIN; 20TH SYMPOSIUM fl NT'L) ON C.umKJUM. pp. 1055-1061. 
t ONE CAMIRA IS PREFERRED. 



rJ 



- Science Applications International Co»par« 



ition- 



POWER REQUIRED FOR EACH FLAME MEASUREMENT (CONT.) 



200 
180 

160 
140 

_ 120 



fe 100 

o 

°- 80 U 



60 

40 

20 





A,B,C,D,E,G,H (170W) 



A,C,D,E,G,H,I (135W) 



A,C.D,E,G,H (120W) fY 



D,E,F,G,H (80W) 



/ 



^D,G,H (30W) 



J L 



10 



20 



30 
Time (sec) 



40 



50 



60 



, — SdencaApplicationslntemationalCon>or 



■Btion- 



ELECTRICAL SYSTEM REQUIREMENTS 

I MEASUREMENT DEVICES: 

• ENVIRONMENTAL TEMPERATURE: 1 TEMPERATURE SENSOR OUTSIDE THE CHAMBER 

• COMBUSTION TEMPERATURE: 9 THERMOCOUPLES ABOVE EACH NOZZLE IS PREFERRED 

• CHAMBER TEMPERATURE: 1 TEMPERATURE SENSOR FASTENED TO THE CHAMBER 

• CHAMBER PRESSURE: 1 PRESSURE TRANSDUCER INSIDE THE CHAMBER 
I RADIOMETER: TBD AND ITS REQUIREMENTS TBD 



CONTROLS: 



I6NIT0R: TBD 

ROTATING-ARH IGNITOR IF THE 2-NOZZLE CONFIGURATION IS USED WITH 

NICHROHE-HIRE IGNITOR 

• ROTATING-ARH THERMOCOUPLES IF THIS OPTION IS USED 

• CAHERA(S): TURN POWER ON AND OFF 

• ALL VALVES. INCLUDING FUEL VALVES. SAMPLING- BOTTLE VALVES, AND 
VENTING VALVE (IF APPLICABLE) 

I INTERFACES: START. STOP. INDICATORS. ETC. 



I DATA COLLECTION: 

I RECORDERS 



I Sdanca Abdications kiiernatianalCai x ntion- 



EXPERIMENT TIMELINE AND SEQUENCE * 



INITIATION OF MEASUREMENTS: 
CAM ERA STARTED AT t = 

I 



FUEL VALVE OPENED AT SPECIFIED 
FLOW RATE (t-0.5 SEC) 



IGNITION FAILS 
TO LIGHT UP THE- 
FLAME IN 5 SEC. 



IGNITION SYSTEM. THERMOCOUPLES 

AND TRANSDUCER MEASUREMENTS 

ARE AUTOMATICALLY STARTED (t-1 SEC) 



L 



MEASUREMENT IS STOPPED BY TURNING 
THE FUEL VALVE OFF. CAMERA. THERMO- 
COUPLE AND TRANSDUCER ARE STOPPED 
WITH A TIME LAG. THE FILM IS 
CHANGED DURING COOLING (IF COOLING IS 
REQUIRED). AFTER PROPER COOLING TIME 
(CONFIRMED BY A THERMOCOUPLE READING 
IN THE CHAMBER). THE NEXT FLAME 
MEASUREMENT IS STARTED. 



r 



•IGNITION IS 
SUCCESSFUL 



THE FLAME EXTINGUISHES (CONFIRMED 
BOTH VISUALLY AND FROM THERMOCOUPLE 
MEASUREMENTS) AFTER SOME TIME BEFORE 
REACHING STEADY STATE 



SAMPLE 
REMOVAL 



MEASUREMENTS ARE 
CONTINUED FOR 
THE REQUIRED TIME 



THE 

FLAME 

REACHES 

STEADY 

STATE 



r BULK OF THE OPERATIONS ARE BY PUSH BUTTONS. 



■ScmromAppSea t ion i kilmiwl kj i m lConiormeon- 



ASTRONAUT INTERFACES 







DISPLAY 
(TBD) 










POWER 
ON/OFF 










ASTRONAUT 
OBSERVATION 




















CONTROL BUTTON TO INITIATE 
EXPERIMENTAL SEQUENCE 





, Sdanottoel*** 3 "* 



Wam»<»on* Corporation- 



SAFETY 



. Sc»»wA«>ficj«foniW^*«tori*Corpor*fafi- 



SAFETY ISSUES 



FUEL CARRIED ON BOARD 

• TRIPLE CONTAINMENT IS REQUIRED 
(PRESSURE VESSELS) 

• AMOUNT CARRIED: TBD 

• IF RELEASED IN SPACELAB OR IN 
COMBUSTION CHAMBER. BUST BE BELOW 
LOWER FLAMMABILITY LIMIT EVEN UNDER 
EXTREME CONDITIONS OF HIGHEST 
TEMPERATURE, EMERGENCY LANDING. 
ETC. 



FLAME IN THE EXPERIMENT CHAMBER 

• EM INTERFERENCE WITH SHUTTLE 
OPERATION IF SPARK IGNITION 

• PORTS FOR THERMOCOUPLES. 
IGNITOR(S). PROBES 

• FUEL TUBING SYSTEM AND VALVES 

I DOUBLE CONTAINMENT 
• VIBRATION IS CRITICAL 
I OTHER FLAMMABLE MATERIALS 
(E.G.. FILM) 

• MATERIALS OF CONSTRUCTION 

• SAMPLING BOTTLES (MUST BE 
EVACUATED BEFORE EXPERIMENT) 



— Som-K* ApplKJtnns htumaHonal Corpaai k xi . 

CREW OPERATIONS* 

• EXPERIMENT IN SPACELAB 
EXPERIMENT ON 

FUEL FLOW, CAMERA (S). IGNITION. 
THERMOCOUPLES. TRANSDUCER ON 
CONFIRM IGNITION 
OBSERVE COMBUSTION 
REMOVE SAMPLES 
FUEL FLOW AND CAMERA OFF 
EXPERIMENT OFF 

CONTINUE TO RECORD TEMP. AND PRESSURE 
APPLY COOLING (FAN) 
EXCHANGE FILM CASSETTES 
REPEAT FOR NEK FLAME 



* 



BULK OF THESE OPERATIONS ARE BY PUSH BUTTON. 



- Scasnoa Applications International Corporation - 



AUTOMATION 

THE FOLLOWING OPERATIONS CAN BE FULLY AUTOMATED IN SPACELAB: 

• EXPERIMENT ON 

• FUEL FLOW. CAMERA(S). IGNITION. THERMOCOUPLES. 
TRANSDUCER ON 

• OPEN AND SHUT SAMPLE-REMOVAL VALVES 

• FUEL FLOW AND CAMERA OFF 

• COOLING FAN ON AND OFF 

THE FOLLOWING NEED DIRECT ASTRONAUT INVOLVEMENT: 

• CONFIRM I6NITI0N 

• OBSERVE COMBUSTION DIRECTLY 

• CHANGE FILM CASSETTES 



SciancaAfvlicstxxvkitwnalionti Corporation- 



FUEL SAFETY CONSIDERATIONS 

THE FOLLOWING REALISTIC AND HYPOTHETICAL SITUATIONS HAVE BEEN STUDIED IN DETAIL: 



PRESSURE RISE IN THE EXPERIMENTAL CHAMBER IF ALL OF THE FUEL IS RELEASED IN THE CHAMBER 

AND THE MIXTURE IS ACCIDENTALLY IGNITED. 

THE ABOVE SITUATION WILL NOT HAPPEN IF THE AMOUNT OF FUEL CARRIED IS NOT SUFFICIENT TO 

MAKE A FLAMMABLE MIXTURE IN THE EXPERIMENTAL CHAMBER (i.e.. BELOW LOWER 

FLAMMABILITY LIMIT. BASED ON THE AMOUNT OF AIR IN CHAMBER). 

ASSUMING THAT INCOMPLETE COMBUSTION OF CH, * l.Wf-CO ♦ 2H 2 OCCURS DURING ALL OF THE 

FLAME EXPERIMENTS. THE MIXTURE OF H r CO AND H 2 REMAINED IN THE EXPERIMENTAL CHAMBER 

SHOULD NOT BE FLAMMABLE. 

ALL OF THE ABOVE SITUATIONS HAVE BEEN STUDIED ASSUMING THAT THE MIXTURE IS NOW RELEASED 
IN THE ENVIRONMENT FROM THE POINT OF VIEW OF CONTAMINATION. A TABLE OF CONTAMINANTS AND 
THEIR MAXIMUM ALLOWED CONCENTRATIONS IS FOLLOWING. 

• • THE CONCLUSION OF THESE STUDIES IS THAT BASICALLY THE FLAMMABILITY 
IS NOT A PROBLEM UNDER THE UNLIKELY OCCURRENCE OF THE EVENTS 
OUTLINED ABOVE. 



I — Scitnca Applications Hrnnational Corporation- 



CONTAMINATION 



Maximum Allowable Concentrations (MACs) of Atmospheric Contaminants In Manned Space- 
craft (both Soacelab and Mld-deckl Ref: NHB 8060.1 B and SPAH (APP.C) 



MATERIAL 


MOL. WT. 


HACS/7-DAY 


ppm 


mg/m 3 


CH„ 

C 3 H 8 

CO 
NO 
NO, 


16. OH 

44.09 
28.01 
30.01 
46.01 


2700 

500 

25 

5 

0.5 


1771 

901.4 
28.6 
6.1 
0.94 


•• For missions up to 7 days. 

' 



• The concentration of each contaminant In the spacecraft 
atmosphere must not exceed Its MAC value, 

• Group limit concept for a group of contaminants: 

E cj/MACj - T< 1 

where cj • concentration of contaminant 1. 

In addition, the Interaction of contaminant groups that 

have potential additive effects must satisfy the relation 

2 T < 1. 

J J 



■ScmK*Ap(i1k*bansittm imtAj i m/ Ccrpormlian- 



RELEASE OF CONTAMINANTS /N SPACEIAB 



IF THE EXPERIMENT IS CONDUCTED IN THE SPACELAB, THE FOLLOWING 
AMOUNT(S) OF FUEL(S) ARE THE ALLOWABLE MASSES TO BE CARRIED 
ABOARD THE SPACECRAFT BASED ON THE RELEASE IN ONLY 10-m 3 OF 
THE SPACECRAFT ENVIRONMENT' (ASSUMING 1 a tin AND 298K) USING 
THE CONTAMINATION TABLE FOR MACs/7-DAY: 



CHj, ONLY: m CH ^ - 17.71 gr, X v/v - 0.27Z (NOT FLAMMABLE) 
C 3 H 8 ONLY: m^ . 9<01 grj z y/v . 005X (NQT FUWWBLE) 



BOTH Cfy AND CjHg CARRIED ABOARD: 



LH H ♦ C 5 H 8 - 1. 
17.71 9.01 



•PRESSURIZED VOLUMES IN SPACELAB: SHORT MODULE (WITH 
SHORTEST TUNNEL - 38.11 m 3 , LONG MODULE (WITH LONGEST 
TUNNEL) - 77.82 m 3 . 



I — Sewno»Aopfc» tou li>erii«< Mi a/Corpo»« to i- 



MAX/MUM CO PRODUCTION 



ASSUMPTION: 



CH 4 + 1.5 2 — CO + 2H 2 
C 3 H 8 ♦ 3.5 2 — 3C0 + l*H 2 \ 

"CHj. " "WW n C,H. " "tWc 



j,' v.nj, v. 3 n g x^ng ^jHg 



n CO " n CH * 3n C,H 8 ' n H ? ' ^CH. * ""c 



L 3"8 



3"8 



n M>' 1,5n CH h + 3 ' 5n C,H. 



3 n 8 



"total, 1 " n air' n total, f " n alr ~ n 2 ,b + n C0 +n H 2 
"total, f ""air " (1 ' 5n CH<. * 3 - 5n C»H. } + (n CH a + 3n C,H a } + ^CH, + ""chJ 



-3 n 8 
■"alr* 1 - 5 "^ + 3 ' 5n C3Hg 



3 n 8 



"3 n 8 



0/ v 



/ V v co . l£P_ 



'CO 



( n CH<, * »W * 10 ° 



v total "total, f n alr + 1,5n CHj, * 3 - 5n C 3 H 8 



■d 



Sdanc»Appfc»fafBtit«»ii«< » or M /C u rpof«t > on- 



POST-FLIGHT DATA ANALYSIS 



i — SaaneaAp&KxtJorati t mi i a Oo nalC u nMxa ti an- 



POST- f LIGHT DATA ANALYSIS 



• EXPERIMENTAL 

• FILMS WILL BE DEVELOPED AND ANALYZED TO MEASURE FLAHE DEVELOPMENT 
AND OTHER PROCESSES 

I TEMPERATURES WILL BE DEDUCED FROM THERMOCOUPLE OUTPUT RECORDS 

• RECORDS OF CHAMBER PRESSURE WILL BE EVALUATED 

I SPACECRAFT ACCELERATION RECORDS AND 6-JITTER DURIN6 THE FLIGHT WILL 
BE STUDIED; PERTURBATIONS WILL BE EXPLAINED BY THE ANALYSIS AND 
INTERPRETATION OF FLIGHT DATA AND FLAME BEHAVIOR 

• REMOVED SAMPLES WILL BE ANALYZED TO OBTAIN INFORMATION IN RELATION 
TO DISTRIBUTIONS OF HjO. 2 . CO. C0 2 . N 2 « HYDROCARBONS. ETC. 

I RESULTS OF RADIOHETRY WILL BE RELATED TO SOOT FORMATION AND FLAME 
TEMPERATURE 



I Soerc»AfipiKMomkitmim( KM iaiCcrjxntio r 

POST-FUGHT DATA ANALYSIS (CO NT./ 

• THEORETICAL 

I THEORETICAL PREDICTIONS WILL BE USED TO SUPPORT THE ANALYSIS AND 

INTERPRETATION OF THE DATA BASE DEVELOPMENT FROM THESE EXPERIMENTS 
• THEORETICAL MODELS WILL BE IMPROVED USING THE RESULTS OF 

• TEMPERATURE MEASUREMENTS 

• SPECIES MEASUREMENTS 

• CHAMBER PRESSURE RECORDS 

• ACCELERATION RECORDS 

• GLOBAL CHARACTERISTICS (E.G., FLAME SHAPE AND LUMINOSITY) 
I CHEMISTRY AND FLUID DYNAMICS OF THE PROCESS WILL BENEFIT FROM THE 

INFORMATION ON THERMAL AND CHEMICAL FIELDS 
I COMBINATION OF EXPERIMENTAL DATA AND PREDICTIONS FACILITATES MODEL 
VALIDATION IN RELATION TO IGNITION. FLAME PROPAGATION. KINETICS. 
RADIATION. AND SCALING CRITERIA 



I — S-«no>AAp to to » * i tei i _ ^ irfCarp»-*an- 



CODR APPENDIX 
(CALCULATIONS) 



r 



Sonne* Applicaboni IrfmaOonal Corpora** - 



VOLUME Of THE EXPERIMENTAL CHAMBER 










HERE, n- NO. OF MLES, T- MASS FLOW RATE, m - MASS, t - EXPERIMENT TIME, II - MOLECULAR WEIGHT, 
P- DENSITY, V - VOLUME FLOW RATE, P - PRESSURE, ft ■ UNIVERSAL GAS CONSTANT, T - TEMPERATURE, 
F - FUEL, - OXIDIZER, P' - C0 2 , P" - H 2 0, AND i,J. and k - STOICHIOMETRIC COEFFICINETS FOR 
COMPLETE COMBUSTION; SUBSCRIPTS: F - FUEL, OXIDIZER, AND 1 - INITIAL CONDITION, f - FINAL 
CONDITION, b- BURNED, AND I - INERT (NITROGEN). 

A * ?U|1PT 1 0NS ' A ) BURNING OF FUEL HAS NEGLIGIBLE EFFECT ON TOTAL NO. OF MOLES. 
B) CONSTANT T AND P IN THE CHAMBER DURING BURNING. 



\Ei. -(*»') *"£*/* 



ScmncmAfif>llc*Oonatttmim( un t/Cuiv c iwtk3n- 



VOLUME Of THE EXPERIMENTAL CHAMBER (CONTB) 

Charter Envlrcrroit: 90X Nj end 10X Oj < ♦ » 9), T - 29810 
P - 1/2 atm or 1 otm 
X ■ no. of moles of O2 bumed/lnltlal no. of roles of Oj 



Experiment tine • «C sec 
Fuel - CjHg (1 - 5) 




6 8 ID 12 

Fuel Flo* Rote, v p (airVsec)— • 



14 16 



0.a^h^^^& * - — ^ *■' - ^ a- ■ ■ if ■ ■ f ^ ■ ■ ■ ■ af ■ ■ 

— ■»5Cwnc*^flpii C JDoniWwTi^pontfCorpor»Don- 



VOLUME Of THE EXPERIMENTAL CHAMBER (COHTV) 

amber Envlrorment = Air (73X N^ crxl 2W C^. * - 3.76), T » 298JO 

P = 1/2 atm or 1 aun 
X = no. of moles of 02 bumed/lnitlal no. of moles of O2 





0.6 




Experiment time 


= <J0 


sec 


X- 


■0.01 










Fuel 


- CH^ (1 = 


2) 












0.5 


















t 


0M 
0.3 


» 














-0.Q2 


s» 


0.2 
0.1 
















.-0.03 

-0.05 
10.07 




















—0.10 



2 4 6 8 10 12 M 16 

Fuel Flow Rote, V F (anVsec) — 



18 



I — Science AfipticmtionskitmnmionalCorporwIion- 



VOLUME OF THE EXPERIMENTAL CHAMBER (COHW) 

Qxrter Envlrortnent: 90X Nj and 101 O2 <*- 9), T ■ 298K, 
P - 1/2 atm or 1 aim 
X s no. of moles of 05 bumed/lnltlal no. of moles of 05 





1.2 




Experiment time 


■40 


sec 


X- 


0.01 








Fuel 


"CHg (1 - 


2) 










1.0 
















t 


0.8 
0.6 














^0.02 


> 


0,1 
0.2 


■ 










1 — 1— 1 1 1 ■ ■ 


^„^.0.03 
0.05 

r^To.07 

^^0.10 



6 8 10 12 M 16 18 

Fuel Flow Rote, V p (an'/sec)— 



r 



Scene* Applications 



International Corporation - 



VOLUME OF THE EXPERIMEHTAL CHAMBER (CONW) 

Outer Environ^ = Air (791 ^ crd 217. 0? * ' 3.76), T = 298K 
p = 1/2 atm or 1 atm 
X = no. of roles of O2 bumed/lnltlal no. of roles of 02 



Experiment time = 40 sec 
FibI = C 3 Hg (1 = 5) 




2 



6 8 10 12 
Fuel Flow Rote, V p (cmVsec) — 



14 16 



Sdenoa Applications Hnmatknal Corporation 



FIRST FUGHT 



I SIX FLAMES. 

• CHAMBER PRESSURE - 1 atm. 

• ay AIR FLAMES. 

• TWO NOZZLES. 

| 40 SECONDS PER FLAME EXPERIMENT TIME. 

• EXPERIMENT CONDUCTED IN SPACELAB OR MSL. 



1 



I SaenotA^jcaboratttmalionalCorponUK 





COHVinONS; FIRST FUGNT 




i 








FU*E A 


FLtfESB 


'. 


C0W1ENTS 




Al 


A2 


A3 


Bl 


B2 


E3 


FUEL 


CH^ 


CH„ 


Z\ 


CH a 


CH q 


CH„ 


— 


~ '^ — ™" 


OXIDIZER 


AIR 


AIR 


AIR 


AIR 


AIR 


AIR 




NOZZLE RADIUS (cm) 


0.051 


0,051 


0,051 


0.0825 


0.0825 


0,0825 


^___ 




INITIAL CHAMBER 
PRESSURE (atm) 


1 


1 


1 


1 


1 


1 







INITIAL CHAMBER 
TEMPERATURE (K) 


295 


295 


295 


295 


295 


295 


— 






EXPERIMENT TIME (SEC) 


10 


10 


10 


10 


10 


10 


— 




FUEL FLOW RATE 
(cm 3 /sec) 


0,75 


1.60 


2.80 


1.00 


2.90 


1.55 


■ — 






FUEL VELOCITY 
(cm/sec) 


91.8 


195.8 


312.7 


16.8 


135.6 


212.8 


— 






FUEL REYNOLDS NUMBER 


28.1 


60.6 


106.0 


23.1 


67.9 


106.5 


— 






— ScmncmAfiplicttnratitmitMi u nt/Cap o im iu n- 





CONDITIONS 


FIRSTFUGNT 


(CONT 

FLAMES B 


V 








FLAMES A 






2 


COMMENTS 




Al 


A2 


A3 


Bl 


B2 


B3 


CHAMBER VOLUME (m 3 ) 


0.058 


0.058 


0.058 


0.058 


0.058 


0.058 




SELECTED BASE] 

ON = io: o 2 

CONSUMPTION 
AT THE END OF 
THE EXPERIMEN' 


X 2 BURNED, X 


0.5 


1.1 


1.9 


0.7 


2.0 


3.0 


-9.2 








6AS-TEMPERATURE 
RISE, AT (K> 


21 


15 


77 


29 


82 


122 


— 


USJNG THE PRE- 
VIOUSLY-DERI- 
VED EQUATION. 


PRESSURE RISE, 
AP (otm) 


0.07 


0.15 


0.26 


0.10 


0.28 


0.12 




USING THE PRE- 
VIOUSLY-DERI- 
VED EQUATION. 


HEAT RELEASE (kJ) 


1.0 


2.1 


3.6 


1.3 


3.8 


5.9 


17.7 


JSING THE PRE- 
/IOUSLY-DERI- 
/ED EQUATION. | 



. Science Applications International Corporation- 

CONDITIONS; 
HRSTHIGHT 
(CONTd) 



MASS OF FUEL (gr) 



MASS OF H 2 (gr) 



MASS OF ALUMINUM 
CHAMBER (kg); HEIGHT/ 
DIAMETER - 2 



MAXIMUM ALUMINUM 
TEMPERATURE RISE (K) 



FLAMES A 



Al 



0.020 



A2 



0.043 



A3 



0.074 



FLAMES B 



Bl 



0.027 



0.045 



0.097 



0.167 



0.061 



B2 



0.077 



B3 



0.121 



0.362 



COMMENTS 



0.173 



37.84 



37.84 



37.84 



37.84 



0.03 



0.272 



37.84 



0.07 0.11 



0.04 



0.12 



0.815 



37.84 



0.18 



BASED ON 
STOICHIOMETRIC 
lOMBUSTION 
<ITH C0 o AND 
\ 2 AS PRODUCT 



jased on a 
;ylindrical 
hamber with 

L.5 cm THICK- 
tESS. 



ASSUMING THAT 
ALL OF THE 
HEAT RELEASE 
IS SUDDENLY 
TRANSFERRED 
TO THE METAL. 



Somnca Applications kilemational Corporation 



CONDITIONS; FIRST FLIGHT (CONTD) 



MAX. CO PRODUCED (Z Vv) 



FLAMES A 



Al 



0.05X 



A2 



0.11% 



A3 



FLAMES B 



Bl 



0.19X 



0.07X 



B2 



0.20X 



CH„ RELEASED IN THE 
CHAMBER (X v /v) 



0.05X 



0.11X 



0.19X 



ORIGINAL MASS OF AIR IN 
CHAMBER (gr) 



0.07X 



B3 



0.32X 



0.94X 



COMMENTS 



BASED ON CHj,+ 
1.5 2 -CO + 
2H 2 0. NOTE 
THAT LOWER 
FLAMMABILITY 
LIMIT OF CO + 
H 2 IN AIR - 
12.5X v /v. 



0.20X 



0.32X 



0.94X 



58.88 



NOT FLAMMABLE 
BECAUSE LOWER 
FLAMMABILITY 
LIMIT IS 5.3X 
AT ATMOSPHERIC 
P AND T. 



3ASED ON 21X 
) 2 , 79X N 2 AT 
1 atm AND 295K 



I Soenc* Applications Hematanat Ccrpora ti ur i 



CONDITIONS; FIRST FLIGHT (CONTD) 





FLAMES A 


FLAMES B 


s 


COMMENTS 




Al 


A2 


A3 


Bl 


B2 


B3 


MINIMUM CHAMBER VOLUME 
IF ONE CHAMBER PER 
FLAME <m 3 ) 


0.0029 


0.0061 


0.0107 


0,0038 


0.0110 


0.0173 




BASED ON 10: 
OXYGEN CON- 
SUMPTION FOR 




















EACH FLAME. 


OCCUPIED VOLUME IN 


















THE FUEL TANK (cm 3 ) 


3.0 


6.5 


11.2 


1.1 


11.7 


18.1 


55 


BASED ON THE 
SUGGESTED DE- 
SIGN OF FUEL 
TANK, P=10 
atm. 



— Sdmx» Application kXmwtiond Corporation- 



MOTION INDUCED IN THE CHAMBER DUE TO JET MOMENTUM 



"Jet m Jet l test ' m alr u alr 



u alr " m Jet u Jet /m alr 



For Chamber " °- 058 £• m alr " W- 88 ^ 



u Jet , cm/s 



m Jet' gr 



u air, induced' an/s 



Al 



91.8 



0.02 



0.03 



Flames A 



A2 



195.8 



0.013 



0.12 



A3 



312.7 



0.071 



0.37 



Bl 



0.027 



0.02 



Flames B 



B2 



16.8 135.6 



0.077 



B3 



212.8 



0.121 



0.15 



0.37 



SfjgnotApplk^dcrahtamationalCcnxiralion- 



KHB 1700,7 
(Nov. 1982) 

JSC-21000-PIP-33- 
(Baslc, June 1985) 

^Feb. 198.) 
NHB 1700. 7A 



HB-S1 -09-84-1100 
(Draft, Ser, 1984) 

NHB8060.1B 
(Sep. 1981) 

JSC-11123 

(Change 1, Sen.1978) 

JSC-13830A 
(May 1983) 



JSC-18798 



1983) 



ICD 2-1 M0Q1 
(March 1984; 
Errata: Dec. 1984) 

JSC-140B4 
(Sep. 1983) 

JSC-16535 _ 
(Rev. B: Sen. 1982) 



REFERENCE DOCUMENTS 

-Space Transportation System: Payload Ground Safety Handbook" 

"Shuttle/Pcyload Standard Integration Plan for Snail Payload 
Accarodatlan" 
-Principal Investigator's Guide to STS Materials Science Experiments" 

"Safety Policy and Retirements for Payloads Using the Space 
Transporatlcn System (STS)" 

"MSL Users Handbook" 

"Flamnblllty, Odor, aid Offgasslng Retirements end Test Procedures 
fornaterlals lnEhvlrarants that Support Cartxistlon" 

"Space Transportation System Payload Safety Guidelines Hcnfccok" 
"inrlercntotlon Procedure for STS Payloads System Safety Requirements" 
"Interpretation of STS Payload Safety Requirements" 
"Orblter Mlddeck/Poyload Standard Interfaces Control Docunent" 

"Shuttle/Paylcad Standard Integration Plan for Mlddeck-Type Payloads" 

"Orblter Mlddeck Payload Provisions Honcbookj Experiments and Cceraticns- 
Support Division" 



-ScimioeAfvhcMt>orat*mt » t i onatCun M iwt i on- 



REFERENCE DOCUMENTS (CONTINUED) 

• jSC-20052 "Space Shuttle Payload: Design and Development" 

Vol. 1 (Rev. B, Feb. 1985): Executive Overview, 

Vol. 2 (Rev. B, Feb. 1985): Avionics interferes and Requirements, 

vol. 3 (Rev. B, Feb. 1985): Flight Design Guidelines and Reaulrenents, 

Vol. 4 (Rev. L Sep. 1984): Kennedy Space Center Operations, 

Vol. 5 (Rev. B, Feb. 1985): Management Dcamntatlon, 

Vol. 6 (Rev. B, Feb. 1985): Mission Operations Interfaces, 

Vol. 7 (Rev. B, Feb. 1985): Safety Guidelines and ReaulrwHits, 

Vol. 8 (Rev. B, Feb. 1985): Structural/Mecfailcol Interfaces and Reoulreients, 

Vol. 9 (Rev. L Sen. 1984): Thermal Interfaces and Requlranents. 
•Shuttle/Poyload Integration Activities Plan" 



JSC-143S 
JSC-SE-R-COOffi 

JSC-14095 
JSC-07700 
MSFC-SPK-522 
JSC-08962 



"General Specifications MSA/JSC Requirements for Materials 
and Processes" 

"Requirements for the Orblter Crew Canparfflent" 

-voIuib XIV Mission integration Control Bard Configuration" 

"Design Criteria fcr Controlling Stress Corrosion Crocking" 

•Coronation of VCM Data of Nometolllc Materials" 



•Effects of Buoyancy on Gas-Jet Diffusion Flames: 
Experiment and Theory" 



R. B. Edelman and M. Y. Bahadori 



^ t fl A«tr Q n aU tica. VQl. 13, No. 11/12, pp. 681-688, 1986 



-Effects of Buovancv on Gas-Jet Diffusion Flames: Experiment and Theory," by Edelman 
and Bahadori Snot be included in this report, but may be found in Acta Astronautica . 



Section 5* 
•LAMINAR DIFFUSION FLAMES UNDER MICRO-GRAVITY CONDITIONS" 



R. B. Edelman, M. Y. Bahadori, S. L. Olson, and D. P. Stocker 



Paper AIAA-88-0645, AIAA 26th Aerospace Sciences Meeting, Reno, 

Nevada, January 1988 



A revised version of this paper has been published in the AIAA Journal (see Section 9). 




,010 



M<> 




AIAA-88-0645 

Laminar Diffusion Flames Under 

Micro-gravity Conditions 

R. B. Edelman and M. Y. Bahadori, 
Science Applications International 
Corp., Chatsworth, CA; and S. L. 
Olson and D. P. Stocker, National 
Aeronautics and Space 
Administration, Lewis Research 
Center, Cleveland, OH 



AIAA 26th Aerospace Sciences Meeting 

January 11-14, 1988/Reno, Nevada 



For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics 
370 L'Enfant Promenade, S.W., Washington, D.C. 20024 



Haper AIAA 88-0645 



LAMINAR DIFFUSION FLAMES UNDER MICRO-GRAVITY CONDITIONS 

Raymond B. Edelman* and M. Yousef Bahadorl* 

Science Applications International Corporation 

Combustion Science and Advanced Technology Department 

Chatsworth, California 



AIAA 26th Aerospace 

Sciences Meeting, 
January 12-15, 1988, 
Reno, Nevada 



and 



Sandra L. Olson* and Dennis P. Stocker* 

National Aeronautics and Space Administration 

Lewis Research Center 

Cleveland, Ohio 



Abstract 

This paper presents the results of an on-going 
experimental and theoretical effort for the study 
of laminar gas-jet diffusion flames under micro- 
gravity conditions. Preliminary results of igni- 
tion studies of methane and propane flames In the 
2.2-second NASA-Lewis Drop Tower (gravity level 
<10" 5 g) are presented. Also, theoretical analyses 
leading to the definition of experiments to be 
conducted In the NASA-Lewis 5. 18- second Zero- 
Gravity Facility (gravity level <10"»g) and the 
KC-135 airplane (free-floating) are described. 
Experimental and theoretical studies show that 
under reduced-gravity conditions, laminar diffusion 
flames are strongly affected by the combined 
effects of kinetics, radiation and transient phen- 
omena (I.e., flame ignition, stabilization and 
extinction). The results of calculations for 
flames in zero-gravity are presented and ignition 
of a family of flames in micro-gravity which reach 
a near-steady state in the 2.2-second available 
time is reported and analyzed. Observations on the 
distinct nature of the flame color and luminosity 
in micro-gravity are described and the flame behav- 
ior under transient, high-deceleration rates is 
discussed. The design of the test hardware along 
with the description of the types of measurements 
and instrumentation being used in the continuation 
of this work are briefly discussed. 

1. Introduction 

Laminar gas-jet diffusion flames under micro- 
gravity conditions have shown distinct character- 
istics relative to the flames in normal gravity. 1 * 9 
Aside from the primary goal of understanding the 
behavior of fires 1n spacecraft environments. 
Isolation of the effects of buoyancy aids the 
analysis of the coupled physico-chemical phenomena 
of radiation, kinetics, convection, and diffusion 
for flames in normal-gravity conditions. As a 
result, the modeling efforts can benefit from the 
removal of one coupled constraint (i.e., buoyant 
force) from the formulation of the problem. Exten- 
sive studies of micro-gravity gas-jet diffusion 
flames in the past -■' have paved the way for a 
comprehensive, theoretical and experimental re- 
search effort, of which, the present paper is the 
preliminary outcome. 

The background and results of the past efforts 
on laminar flames In micro-gravity environments are 
presented in Sec. 2. Section 3 gives a description 



•Assistant Vice President, Director; Member AIAA 
{Senior staff Scientist; Member AIAA 
♦NASA Project Scientist 
■NASA Experiment Manager 



Tkb mmt k tt H mtt t «-*k «f tta U'.J. 

ax mtf" » i f | i % » pmiid ii l> (H HUM 1 



of the apparatus. In Sec. 4, the experimental 
procedure Is described. Section S presents the 
results and data comparisons, followed by the 
conclusions and description of the future efforts 
In Sec. 6. 

2. Background 

Experimental and theoretical efforts on the 
study of gas-jet diffusion flames 1n micro-gravity 
environments are described and analyzed in a re- 
cent publication. 9 Assessment of the existing 
data obtained under reduced-gravity conditions has 
shown that further research is required before 
reliable predictions of Ignition, stabilization 
and propagation of the flames under micro-gravity 
conditions can be made. In this section, a brief 
description of the studies conducted toward the 
understanding of the behavior of laminar diffusion 
flames 1n reduced-gravity environments is present- 
ed. 

Experimental Efforts 

Studies of gas-Jet diffusion flames under 
micro-gravity conditions have been conducted in 
the 2.2-second NASA-Lewis Drop Tower (<10" 5 g). 
Methane, hydrogen, propylene, and ethylene flames 
have been extensively studied. J" 8 In these stud- 
ies, the flames were Ignited in normal gravity, 
allowed to reach steady state, and then the exper- 
iment package was dropped inside a falling drag 
shield. High-speed movies of the flames have 
shown that during the transitory period of adjust- 
ment from normal-gravity to micro-gravity, a sudden 
decrease in the flame height occurred within -0.05 
seconds. After the decrease of the flame height 
to a minimum. I.e., 0.6 < h„j n /hi_„ < 1.0 (depend- 
ing on the jet Reynolds number), tne flame height 
began to Increase with time, resulting 1n extinc- 
tion with 1.0 < h ex /hj.g < 1.3, reaching an appar- 
ent steady state with 1.4 < ho_ 4 /hi_g < 1.7, or 
remaining 1n a transitory stately tne time the 
package came to rest in a sand box at the bottom 
of the tower. Compared to the flames 1n normal- 
gravity, larger, rather globular flames are obser- 
ved 1n micro-gravity which are due to the absence 
of buoyant force, making diffusion the major mech- 
anism of transport. The orange-reddish appearance 
of these hydrocarbon flames *" 8 is due to the 
significant amounts of solid carbon formed by 
pyrolysis in the hot fuel-rich portion of the 
flames. 

The observed flame behavior 1n micro-gravity, 
i.e., reaching an apparent steady state, being in 
a transient mode, or extinguishment, has been 
attributed to the influence of the jet Reynolds 



1 



?*<* e 



c*^V^ 



number. 1.t.. the combined effects of nozzle size, 
Jetveiocity and fuel type. Figure 1 shows the 
behavior of two methane-a1r flames L* n ■ ero- 
oravlty. I.e.. an extinguished flame [Fig. 1(a)] 
2nd a flame which apparently reached steady state 
ma Kb)] within the 2.2 seconds of the drop. 
The data suggest that for a certain range of fuel 
velocity, a transient period occurs between a near- 
steady state and flame extinguishment. However, 
the results of the present study (see the section 
on experimental results) have shown that a »*Jo r J ty 
of the flames which were in a transitory state In 
the previous studies were indeed ignited and reach- 
ed a near-steady state once the flame was ignited 
in micro-gravity. The source of the difference 
1n the behavior of the flame is the method of 
ignition. If the flame 1s ignited in normal- 
gravity and then dropped, the transitional behavior 
is due to the accumulation of the hot products of 
combustion In the flame region which results from 
the sudden loss of the buoyancy-driven convect ve 
flow. This can enhance soot formation, radiation 
cooling and the onset of a ^emical kinetic llmta- 
tion on the heat release process. The combination 
of these effects leads to either quenching or a 
slow readjustment to a steady-state flame, depend- 
ing on the jet momentum and the fuel mass-flow 
rate. Even a flame which reaches an «PP«™"J 
steady state [Fig. 1(b)] shows some oscillations at 
the final stages of the drop which suggests that 
the flame may be 1n transition from a transient 
state to extinguishment. 



Theoretical Studies 

A mathematical model has been developed 6 > 7 
for the study of laminar gas-jet diffusion flames 
under arbitrary gravitational accelerations based 
on the parabolic form of the equations of motion. 
The model Includes the effects of Inertia, vlscos- , 
1ty. diffusion, radiation, and chemical reactions. 
The differencing scheme is an explicit finite- 
difference technique. The chemistry «ssu"f d 1s 
that of shifting equilibrium, and both Flcklan and 
multl -component diffusion are considered. Figure 
2 shows a comparison between the predicted and 
measured flame heights under both normal-gravity 
and zero-gravity conditions. Excellent agreement 
was obtained for flame height predictions (Fig. Z). 
However, It was shown 6.7 that accurate flame shape 
predictions require an accurate prediction of tne 
relative rates of the molecular transport of spec- 
ies, momentum and energy. 



•j. i 



1m 



(•) 



Fwl Hob riU*» 

• I.t ct/» J 

■suit rW1« i 

• 0.1M m i 



Extinction 




!l.!" 



r 



FmI r\m rata 

• l.l etyt 

Mult raflM 

• «.M1 <a 



't.J 



££ 



i.) 

4. J 



2lf«-€ 



S.J 
IK 



I.? 



»•• 



r-Y.; 



»»t. nmH 

Kit. I«r»-l \iM<» 



ff^yTlv 



Fig. 1 



Methane-a1r flame lengths as a function 
of time in micro-gravity for (a) an 
extinguished flame. l ind (b) a flame 
which reached an apparent steady state 
at the end of the drop. * 

The sensitivity of the extinguishment P">CS» 
to convectlve transport has been demonstrated; 
low forced-air velocities (=»10 cm/s) In co-ax1al 
jets of methane-a1r diffusion flames In ■Icrc- 
gravlty were sufficient to sustain combustion for 
methane flow rates that otherwise resulted n ex- 
tinguished flames. The Ignition *&">'<*'*'*•' 
Igniting the flame 1n normal-gravity and JhM J/f: 
ping the apparatus, the test time Imitation of 2.2 
secSnds. and the lack of quantitative data have so 
far been the weaknesses 1n achieving a more funda- 
mental and quantitative understanding of gas-jet 
diffusion flames In micro-gravity environments. 
However, this past experimental and theoretical 
work 1-4 has provided the basis for the focused, 
quantitative approach described 1n this paper. 




tip. 


lm«nlal do'a 


Nof mol - 


r 


'o. e» 


O 




05' 


* 




0(25 







om 







' M 


© 




i» 


» 




**i 



Son* «r"ieo« •»"°'« 
imre-t 



100 



"•• 



Fig. 2 



Comparisons between the theoretical 
predictions and experimental results 
for non-dimensional flame heights vs. 
jet Reynolds numbers; methane-air 
flames. *•' 



Flame ignition and development toward steady 
state 1n zero gravity requires a time-dependent 
analysis. An analytical model was developed * for 
the shapes and heights of zero-g, laminar, cylin- 
drical^ symmetric, transient diffusion flames with 
the effects of axial diffusion included. The ana- 
lysis shows that the theoretical steady-state flame 
lengths are approached slowly. However, the theory 
did not Include the effects of radiation and kine- 
tics. As a result, the predicted approach to 
steady state may be faster 1f these effects were 
taken Into account. Figure 3 shows a comparison 
between the analytical and experimental flame 
heights under micro-gravity conditions, where D is 
the binary diffusion coefficient of methane 1n 
nitrogen, U 1s the fuel velocity, R* 1s the nozzle 
radius, h s . 1s the theoretical steady-state flame 
height, and a value of D ■ 1 c*2/s corresponds to 
a temperature of 750 K. 

Flames can be strongly affected by the level 
of gravitational acceleration acting upon them. 
The change In the flame behavior due to the convec- 
tlve effects can be seen In Fig. 4, ""ere the 
centerllne velocity as a function of axial distance 
for different gravitational levels s shown. in 
Flo 4 U„ ■ 93.5 cm/s 1s the velocity of the metn- 
J2'jJt and L - 50 « 1s the normalizing axial dis- 
tance, with the nozzle radius - 0.0825 cm. 







5 



• «(»»* tA«lfflC0l I 

J 1 . ■ 



10 



1 5 



19 



Fig. 3 Analytical and experimental varlatl 

of f ante heights with time under micro 
gravity conditions. 9 



ons 




Fig. 4 Centerline velocity vs. axial distance 
along the jet for a methane-air flame 
as a function of gravitational level. 9 

Residual accelerations (l. e . a -iitt#r) ran 
cause fluctuations In flame neigh .* ee c Or- 
ations exist in the KC-135 airplane and In the 
spacecraft during maneuvering. Using an order-of 

with the aid of the steady-state model 6. ? alreadv 
described, an approximation for the effects of fre- 
quency of oscillations in the g-levet on the flaw 
height can be obtained. Plguri 5 shows a £su Hof 
such analysis. The calculation was performed for I 
»ethane-a1r flame with a jet velocufof gTs 2/s 
nozzle radius of 0.0825 cm. gravity level of 1?V 
and a 31 allowable fluctuation in f lime Je?gh ?n 
F'9- 5. any cognation of amplitude ft) and fre- 

?hT^l L U) ° f J1tt ' r ^ r " uU$ <"• Point above 
the solid curve must be damped. 

„ nt „"* Jt «??y-state model described previously Is 
not only applicable to the description of the fun - 
damen al processes, but also to the design of tte 

Ilr*t£ f^H J M t1nfl d,fferent 'utl-flo. rates In 
air, the flamnable region which the spark must oass 
through for successful Ignition of the flaw in 

JSi;f: i o} t ? u , h iden r f , ,ed - "'^ * «»?•" 

example of such a calculation. 

f} — l mt ri Ji* 7 ' t r ip «:» tu r e distributions above a 
JlEltJ # rr? rent " 1al 1ocat ^" «re shown. 
Results of this nature have aided in identifying 



the proper locations for the instrumentation (. a 
thermocouples and sampling probes). ( *' 



I 10*' 



W J • 




u,-«.5 an 

Lf^.Hl ft • S.H a 
(AL, ello-«i/W - « 



-J -I -i 



Fig. 5 Effect of residual accelerations 
and the dampening requirement. 




F1g. 6 Flammability regions for cold jets 
of methane in air. 



A: 


i ' 10. i m 


• : 


i ■ 12. » ca 


C: 


i ■ IS. 2 at 


0: 


i ■ 17.9 cm 


£: 


> * 20. S o 


F: 


i ■ 23.2 a 




CM 4 /4fr n«B 

4 " 300 cc/»1n 

I.* 0.OS1 cm 

Flat iMtgM • t.7 ca 



Fig. 7 Temperature distributions above a 
methane-air flame in micro-gravity. 

The transient effects described In this sec- 
tion, I.e., Ignition, flame development toward 
steady state, extinction, sensitivity to the 
effects of chemical kinetics, and the attended 
radiation as affected by these processes, cannot 



Of ?°°* 



be adequately analyzed without a more complete 
transient. mu1t1 -dimensional analysis of these 
flames. The Intent of the experimental data, 

1 e , temperature, species and radiation measure- 
ments 1s largely to provide support In the formu- 
lation and validation of such a model that is 
under development in this research. 

3. Apparatus 

The experimental studies were conducted in the 

2 2-second NASA-Lewis Drop Tower (Fig. 8). The 
tower consists of a building 6.4 x 6.4 x 30.5 m. 
The drop area in the building is 1.5 x 2.75 x 27 m. 
The 2. 2-second period of micro-gravity was obtained 
by allowing the experiment package to free-fall 
from the eighth floor. The experiment was suspend- 
ed from the ceiling by a highly stressed music 
wire, and then released by pressurization of an air 
cylinder that forced a knife edge into the support 
wire which rested against an anvil. 



tfPERIAtfNT IN 
PttOROP POSITION-; 



CHECK OUT 
AREA— - 




ELEVATOR 



DOWN 



-SAND STORAGE 



«-Kati*AT10N 
CONTAINER 



Fig. 8 2. 2-second drop tower. 

The experiment package was allowed to free- 
fall Inside a protective air-drag shield (Fig. 9). 
The drag shield had a high ratio of weight to fron- 
tal area and a low drag coefficient, resulting in 
a gravity level of less than 10" s g acting on the 
experiment package during the 23-m fall. The alum- 
inum spikes of the drag shield decelerated the 
package in a sand container located at the bottom 
of the drop tower. 

The experiment package and a schematic diagram 
of the flow system are shown 1n F1gs. 10 and 11. 
respectively. The apparatus consists of a combus- 
tion chamber which contains a burner, the lighting 
system, a viewing port for the high-speed camera, 
the spark Ignitor, and ports for filling the 
chamber with oxidizer and venting the burned gases. 
The other components of the system were a metering 
valve, solenoid valve, pressure gauge, pressure 
regulator, fuel bottle, hand valves, and vents. 
The volume of the combustion chamber was approxi- 
mately 40,000 cm 3 . Two fuel nozzles with Inside 
rad11 of 0.051 cm and 0.0825 cm were used In the 
studies. The detailed sketch of the burner and 
dimensions of the tubes are shown in F1g. 12. Tne 



r MUSC wiK 

\SUPPO*1 
\. - ^-»l« RELEASE MECHANISM 

V* ^OUACSMCO 

rr tt tt I I 



-tAS£ ROUNDED TO 
RtDUCE AIR OMC 

O * * U ii 11 

IW KfOW TtST OROR^ ^ v 

(II DURING TEST DROP. 

IC) AFTtR TEST DROP. 



4^0 



Fig. 9 Position of the experiment package 
relative to the drag shield. 




Fig. 10 Experimentpackage: (T) Combustion 
chamber; © Flowmeter; © Viewing 
window partially hidden by strut^4) 
Solenoid valve hidden by strut; (57 
Mixing system; (6) Pressure regulator; 
© Fill line; (8) Control panel; (9) 
Fuel bottle; OT Camera partially 
hidden by regulator and fuel bottle; 
<Q to vent. 




Fig. 11 Schematic diagram of the flow system: 
(f) Experiment chamber; ® 8umer 
Sozzle; © Chamber bulkhead; © 1/8 
1n. o.d. tubing; (?) Metering valve with 
Vernier markings; (£) 1/4 In. o.d. tub- 
ing; (7) Solenoid valve; © Pre«ure 
regulator; ® Hand valve; © Quick 
disconnect; gT> Pressure gi^L.^? 
Fuel bottle; T> Hand valve; © Quick 
disconnect; dj Vents. 



pFvGfc 



£$& r c«5£ 



-1».» 

cat 



u 



111*1 10 

clwmoar 
•uklM«4 



■3) 



Wll — 1|— 



tLkmJ 


OO (em) 


KiaJcm, 


0.102 

one 


0.1*1 
0.22« 


o.oaoa 

0311 



Fig. 12 Sketch of the burner and tube dimensions. 

burner tips Mere cut off and were not machined to 
a sharp edge. The high-speed Mil liken movie camera 
(16m»J mis originally set for a rate of 40 frames 
ptr second. As a result, some of the low-flow rate 
flames were not quite visible 1n the processed 
(Tungsten Eastman Ektachrome Video News) films. 
Visibility was subsequently Improved by reducing 
the framing ra te to 20 frames/sec. 

As previously cited, theoretical calculations 
were used to guide the location of the spark elec- 
trode at 4 mm above the nozzle tip and 4 mm away 
from the nozzle centerllne. The electrode was 
positioned approximately 90° to the centerllne of 
the camera. A ground wire was attached to the 
nozzle, making It the second electrode. A sche- 
matic diagram of the electrical system for the 
Ignltor Is shown In Fig. 13. The spark energy and 
frequency were approximately 18 mj and 4.8 Hz. 
respectively. For all of the tests, the spark- 
Ignition system was left on for the entire test 
time of 3.0 seconds, i.e., 2.2 seconds under micro- 
gravity condition and about 0.8 seconds after 
Impact. The Ignltor occasionally did not spark at 
the burner tip but shorted out elsewhere. However, 
this did not affect the flame behavior. A frame- 
by-frame study of the movies showed no disturbance 
of the flames by the spark during the 3.0 seconds 
of burning. 



7O-150V 



HV Power 
- Supply - 




Spark gap 



3xf 



n 



£> 




♦15V 



Sj M sjaiSs_ 

ioo*» ^Mi .on'i 



'SCR 



1N4O0s4M -L 

4 



jioc*n 



F1g. 13 Electrical system for the Ignltor. 

The metering valve (with Vernier markings) was 
calibrated for 1 psig upstream pressure, and the 
pressure regulator had a 0-5 psig capability. The 
fuel bottle was approximately 500 caH 1n volume, 
filled with either methane or propane. The dry air 
in the combustion chamber was composed of 21J 
oxygen and 79* nitrogen, and had a pressure of 1.0 
t 0.01 atm In the chamber. 



<■ Experimental Procedure 

"ethane and propane flames were ignited under 
both normal-gravity and micro-gravity conditions 
and the flame behavior was recorded during the 3 
seconds of burning. This section describes briefly 
the initial setup, test setup, normal -gravity test 
drop-test preparation, drop-test procedure, post- 
drop procedure, and post-test procedure. 

Initial Setup 

The desired nozzle size was installed in the 
experimental chamber and the spark electrode was 
positioned. After sealing and evacuating the 
chamber, dry air was Introduced 1n the chamber to 
raise the pressure to 1.0 t 0.01 atm. The fuel 
bottle was then evacuated and filled with either 
methane or propane to approximately 2.0 atm The 
fuel pressure regulator was set for a pressure 
difference of 1 psl. and the film was loaded in 
the high-speed camera. 

Test Setup 

The total test time and the desired duration 
for the spark generation were set. The filming 
rate was selected and the fuel flow rate was set 
using the calibrated metering valve. 

Normal -Gravity Testing 

8efore conducting the particular normal- 
gravity test, several tests (not filmed) were con- 
ducted to verify ignition and fine-tune the 
pressure regulator. After successful Ignitions, 
the normal -gravity test was filmed. The lights in 
the room were dimmed and the chamber window and 
camera were covered. This was not necessary in the 
drop tests because the experiment package was in- 
side the drag shield. The master and camera jump- 
ers were Installed and the master switch was turned 
on. By turning on the ready switch, the chamber 
light was automatically turned on. The movie 
camera was started and given approximately 3.0 
seconds for start-up, at which time the master and 
camera Jumpers were quickly removed, causing the 
chamber light to go out, the fuel solenoid valve 
to open and the spark generator to start. After 
the test was completed, the camera stopped, the 
solenoid valve was closed and the spark ignitor 
stopped. The master, ready and camera switches 
were then turned off. The film was advanced (to 
leave a blank space) by Installing the camera 
jumper and turning the camera switch on for a few 
seconds. 

Drop-Test Preparation 

The pressure gauge was removed to prevent un- 
necessary fatigue during the deceleration period. 
The experiment package was placed In the drag 
shield and hoisted to the top of the tower. The 
master and camera jumpers were then connected 
through the drag shield. A pre-drop burn (not 
filmed) was conducted while the package was at the 
top of the tower to ensure that there was enough 
fuel at the burner tip. The switches were then 
turned off and the master Jumper was reconnected. 
The camera Jumper was connected to a remote switch 
so that the connection could be made without dis- 
turbing the package. 






Drop-Test Procedure 

The master, ready and camera switches were 
turned on, resulting 1n the turning on of the 
chamber light. The camera did not start here as 
the Jumper connection was open at the camera Jumper 
switch. The drag shield was then closed and the 
package was steadied as 1t was connected to the 
music wire. The camera jumper switch was turned 
on, and after allowing approximately 3.0 seconds 
for the camera start-up, the drop was initiated. 
This was accomplished by cutting the wire via a 
hydrauHcally powered blade with the help of the 
weight of the package. The jumpers were pulled, 
resulting 1n the start of the experiment. Table 1 
shows the drop sequence. 

Post-Drop Procedure 

The drop package was electrically grounded to 
dissipate any charge built up on penetration Into 
the sand. The package was hoisted back up to the 
preparation area. The master, ready and camera 
switches were turned off. The package was removed 
from the drag shield. The film was advanced to 
leave a blank space on the film between two tests. 
The pressure gauge was reattached and checked for 
any changes In the fuel-bottle pressure. Then, the 
test set-up procedure was repeated for the next 
test. 

Post-Test Procedure 

The excess fuel 1n the fuel bottle was burned 
off and the bottle and experiment chamber were 
evacuated, exhausting the contents outdoors. The 
film was removed from the camera (if necessary) and 
the batteries were recharged (if necessary). 

5. Results 



Table 2 shows the experimental conditions and 
the average steady-state flame lengths, maximum 
flame radii, and approximate time to reach steady 
state for both normal -gravity and micro-gravity 
methane and propane flames. The normal -gravity 
flames were flickering in most cases and the 
steady-state flame heights and maximum rad11 are 
the averages. Figures 14 and 15 show the flame 
lengths as a function of time for methane and pro- 
pane flames 1n micro-gravity, respectively. The 
methane flames at low volume-flow rates were very 
faint. However, they were sufficiently visible to 
Identify the maximum radius, but beyond this 
height, visible radiation could not be observed and 
the flame appeared to be under- ventilated. This Is 



a • 

« 



o o 



0° 



ltd rl !.! 




Sj— I 



lull* 

la«in(ca) 



1 



F»l 11m, 



l.t 
».t 

l.» 
JO 

J.t 

l.t 



a!a ' B ' i ' * ' » ' I ' l ' i ' B 1 » ' « ' *■ ' <•» ' * 



Tint* <MC> - 

Fig. 14 Micro-gravity methane-flame heights. 



u 

10" 



o o o ° o o e 



So ° 
CM rl 12- 



B 8 I 



S 




V* 


Ml»l>) 


f-*i rim 
UtMm*t.) 


• 
* 
■ 
o 

A 



T 

• an 

1 


l.l 
I.* 
•.1 

l.l 
l.t 
t.l 



Fig. 



^14 04 1J l.« 1.0 J.4 31 3U 

15 Micro-gravity propane-flame heights. 



shown by the truncated data given In F1g. 14. The 
planned longer-time experiments together with temp- 
erature and concentration measurements will help to 
clarify the structure of these flames. The higher- 
flow rate methane flames reached near-steady state 
with some fluctuations at the later times. These 
flames with a 0.0825-cm nozzle radius were in a 
transient state 1n the previous studies * where the 
flames were ignited in normal -gravity and then 
dropped. Figure 16 shows the corresponding compar- 
isons. 

The steady-state parabolic model 6 > 7 described 
previously was applied to the zero-gravity flames 



Table 1. Drop Sequence 



Time (sec) 



Event 



-3.0 (Approx.) 




2.2 

3.0 (Approx.) 



Camera (with timing light) starts operating. The chamber light 1s on. 

Experiment package drops. Chamber light goes out. Fuel solenoid valve opens. 

Spark electrode Is activated. 

Fuel Jet Ignites. t t .„ 

Drop package impacts the sand box. Flame may extinguish on Impact, then re-Ignite 

from sparking. 

Experiment ends. Camera stops. Fuel solenoid valve closes. Spark generation 

stops . 



♦Test results have shown that -6. 2-6. 5 seconds is needed for the flame to Ignite due to the frequency 
of -»»4.8 sparks/sec. 



Table 2 Data obtained from micro-gravity and normal-gravity experiments' 



Flame 


Nozzle 




Fuel volume 


Steady-state 


S.S. Maximum 


Approx. time 


Reynolds 
number^ 


number 


G radius, 


Fuel 


flow rate, 
cm^/sec 


flame height, 


flame radius, 


to reach S.S. , 




cm 




cm 


cm 


sec 




1 


0.051 


Methane 


1.0 


1.5<: 


0.67 


-0.6 


38 


2 
3 


1 






2.0 


3.5 


0.70 


0.8 


76 






3.0 


5.1 


0.75 


1.0 


114 


4 


0.0825 






1.0 


2.4C 


1.06 


-1.0 


24 


5 


1 






2.0 


3.7<= 


1.05 


0.9 


48 


6 




3.0 


5.5 


1.11 


1.6 


72 


7 


0.051 


Propane 


0.5 


4.0 


0.90 


1.2 


70 


8 
9 








1.0 


7.2 


1.02 


0.8 


140 


I 






1.5 


11.1 


1.11 


-0.8 


210 


10 


0.0825 






0.5 


3.8 


1.32 


1.2 


44 


11 








1.0 


7.3 


1.53 


1.8 


88 


12 


1 


t 


1.5 


9.9 


1.73 


1.3 


132 


13 


[ 0.051 


Methane 


1.0 


0.9 


0.35 


-0.8 


38 


14 








2.0 


2.0 


0.43 


1.3 


76 


IS 


1 






3.0 


<-0(?) 


0.36 


(?) 


114 


16 


0.0825 






1.0 


0.9 


0.38 


0.6 


24 


17 








2.0 


2.5 


0.60 


0.4 


48 


18 


1 






3.Q 


3.6 


0.61 


-0.3 


72 


19 


0.051 


Propane 


0.5 


2.6 


0.45 


0.4 


70 


20 

21 








1.0 


5.6 


0.50 


0.4 


140 


1 






1.5 


8.0 


0.50 


1.3 


210 


22 


0.0825 






0.5 


3.0 


0.57 


0.4 


44 


23 
24 


1 






1.0 


7.8(?) 


0.50 


(?) 


88 




1.5 


8.0 


0.58 


-0.4 


132 



Question marks indicate uncertainties In the measurements due to the flame flicker. 
°Reynolds number is based on the nozzle radius. 
c These flames were diffuse, faint, and resembling an under-ventilated flame; it could not be concluded 

that they reached steady state (see Fig. 14). 



studied in the present work. The results are shown 
in Fig. 17, where a satisfactory trend in the pre- 
dicted behavior of the flames Is demonstrated. 
However, the propane flames are slightly longer and 
the methane flames shorter than the predicted 
flames. Since soot Is not explicitly accounted for 
In the model, the radiation loss is not accurately 
predicted. It appears that the slight over- 
prediction of the methane flame heights Is due to 
the over-prediction of the gas radiation which 
results 1n slightly cooler, and hence, taller 
flames. On the other hand, propane Is known to 
soot more prolifically than methane, leading to a 
substantially higher radiation loss than is pre- 
dicted by the current model, even though the gas 
radiation may be somewhat over- predicted. 



In micro-gravity flames, the heat release 
mechanism 1s not uniform throughout the flow 
field. 6 Pyrolysis (producing soot) and partially 
oxidized species dominate 1n the near-jet region, 
while the continued heating and increased residence 
time result in the tendency to burn off these 
species downstream of the flow. However, pyrolysis 
and soot formation reduce the temperature level in 
the flame, and with the aid of the increased resi- 
dence time, sooting Is enhanced, which, in turn, 



a 

r • 
J . 



i Trmiwt fl*an (m k»(*t tntlttu) 

1 .... f» *u itallakla 

»•«*' < hu mll*»l« 

I A ItHtMUU n« 
V • Utlaarisaaa flam 




O ..-; 



■wli (Mint, 

% • I.Ml m 



' ? ' 'III 1 1 1 I I ■ ■ ■ • ■ ! 



uo \m mo un 



I. 

3 

r 

•> * 

1 





rr — 




% 


u». 


TkMf? 


Fwl 


<<■) 


A 




«4 


•.Ml 


O 


— 


«■ 


*.mn 





.... 


«y% 


• .Ml 







'*• 


o.otn 





Q. 




ttrm\4% 



Fig. 16 Methane flaaie heights as a function 
of Reynolds number. 



• 1 1 3 

fatl Natl Ft*, fttu 1 l») (f/>) . 

Fig. 17 Predicted and measured flame heights 
1n micro-gravity. 



S-&S5& 



results In enhanced radiation, more cooling, and 
delayed soot burn-off downstream of the flame. 

The observed open tip of the propane flames 
and quenching at the tip (see Fig. 18), giving the 
appearance of an under-ventilated flame 1s due to 
the combined pyrolysis, cooling and soot-formation 
processes. 

In Fig. 18, a frame-by-frame analysis of 
Flame No. 11 of Table 2 Is presented, where the 
flame reaches steady state and remains fairly un- 
iform throughout the drop. The orange- reddish 
color of the flame is due to the soot at cooler 
temperatures than would be expected for typical 
burn-off conditions. In addition, the other colors 
indicate the presence of particular species pre- 
sent in hydrocarbon flames, e.g., blue-violet 
(outer region at the base showing the spectrum of 
burning CO), bluish-green region (showing C2 and 
CH emissions where Cj bands appear green shaded 
toward violet and CH bands appearing violet-blue), 
and highly luminous zone (bright yellow due to 
the continuous spectrum of the thermal emission 
of burning carbon particles) which becomes yellow, 
then orange, and then red and dark red toward the 
boundary of the visible region as the temperature 
of the unburned soot decreases. 

The general Inverse relationship between wave- 
length and temperature 1s consistent with Wein's 
law which implies an Increase In the wavelength 
with reduction in temperature. The burning soot 
particle is yellow, I.e., ~0.58-u« wavelength 
(1370 K < T < 1670 K, where 1670 K is the tempera- 
ture of a white hot body). As the temperature 
decreases, the soot particle becomes orange red 
(—1370 K, ~0.6 urn), then cherry red (~1170 K, 
~0.6-0.65 urn), and finally dull red (~970 K, 
~0. 65-0. 70 pm). These effects are clearly shown 
in the results of F1g. 18. which show the cooling 
by pyrolysis and enhanced radiation due to Increas- 
ed soot formation. 

As a consequence of the systematic study of 
flame color and luminosity, important species 
identification will become possible, and when com- 
bined with species and temperature measurements, a 
data base for the model development is anticipated. 

Frame number 35 of Fig. 18 shows the moment 
at which the package Is decelerating at some high 
rate. The Induced convectlve effects transport 
the hot products of combustion away from the 
nozzle, causing flame extinction within <0.1 
seconds. Re-1gn1t1on can be seen due to the spark 
at a later tine, resulting 1n a behavior observed 
in normal-gravity. I.e., flickering. A study of 
the high-speed movies of this segment of the test 
will reveal the behavior of the flame 1n the tran- 
sient state under rapid changes 1n gravity level, 
and when combined with the transient analysis, will 
serve as a by-product data base for further analy- 
sis. 

6. Conclusions 

A spark Ignltor has been developed which suc- 
cessfully Ignites methane and propane diffusion 
flames In micro-gravity environments. 

In the prior research on gas-jet diffusion 
flames, ignition was accomplished in normal- 
gravity, and the established flame was then sub- 



jected to the micro-gravity condition. This gave 
rise to a transient behavior that was experiment- 
related, resulting 1n the questions of ignitabillty 
and flame development in micro-gravity. The pre- 
sent study shows that those transient flames, in 
most cases (I.e., higher flow rates), reached a 
near-steady state In the 2.2 seconds of the drop. 

The lower-flow rate methane flames (1.0-2.0 
cnw/s) appeared faint, diffuse, and resembling an 
under-ventilated flame. This suggests that longer 
test times and quantitative measurements are needed 
to truly identify the behavior of the flames at 
very low flow rates, and possibly, obtain a limit- 
ing flow rate, below which, the flames extinguish. 

The same argument Is applicable to propane 
flames, where all of the flames reached a near- 
steady state with Intense yellow to orange to red 
toward the boundary of the flame, and an open tip, 
showing under-ventilated behavior. This 1s attrib- 
uted to the effects of radiation and cooling which 
stop the soot burn-off process and result in 
quenching at the flame tip. Although the propane 
flames (0.5-1.5 cm 3 /s) did not quench, there should 
exist a lower limit on the flow rate, below which, 
the propane flames extinguish as well. 

The predicted flame heights using a steady- 
state parabolic model were In good agreement with 
the measured steady-state flame heights. The 
slight over-prediction and under-pred1ct1on for 
methane and propane flames, respectively, is due to 
the over-predictat1on of gas-phase radiation for 
methane flames and under-predlctlon of total radia- 
tion (i.e., soot and gas-phase) for propane flames. 

It 1s also apparent that the flame colors can 

aid 1n the Identification of major species and 

regions of soot burn-off and quenching due to soot 
radiation. 

The planned longer test times provided by the 
5. 18- second Zero-Gravity Facility (<10 _5 g) and the 
KC-135 airplane (free-floating, 10-15 seconds) 
combined with flame visualization and temperature 
and concentration measurements will provide a more 
complete quantitative description of these flames, 
which will be used as a baseline for data compari- 
sons and model development. 

Acknowledgements 

This work Is supported by NASA-Lewis Research 
Center under the PACE (Physics and Chemistry Exper- 
iments In Space) program, contract NAS3-22822. 

The authors would like to express their ap- 
preciation to Professor Irvln Glassman for a 
discussion on the Interpretation of the colors 
observed 1n micro-gravity flames. They also wish 
to thank Mr. Matthew Hart of Purdue University for 
the assembly of the experiment package, Mr. Thomas 
MorMssey for conducting parts of the numerical 
calculations, and Ms. Angelica Ciraulo for the 
graphics and typing the manuscript. 



\S 



c*^*^ 



0* 



?oo* 




Frame-by-frame analysis of micro-gravity 
propane flame No. 11 of Table 2 with noz- 
zle radius ■ 0.0825 cm and volume flow 
rate - 1.0 cm 3 /s. The different observed 
colors are as follows: □ Bright white; 
13 White with little yellow; 02 White 
with more noticeable yellow; Blue; 
Dark blue; S V1o1et-p1nk; gHlxed blue? 
violet; Yellow; g) Orange; ■ Cherry 
red; Q Dull red, fainting; Col or at 
the base, starting with dark at the bot- 
tom, becoming dark blue/pink, then pink/ 
blue, then pink/orange, then orange/ 
yellow, and then yellow/white toward the 
center of the flame; E Bright white 
spark with bright blue at its boundary. 
Scale: H :i». Note the asymmetry 
caused by the electrode. 



^Vo^ 






(Continued) 




References 

Cochran, T.H. and Maslca. M.J., "Effects of 
gravity on Laminar Gas Jet Diffusion Flames,' 
NASA TN 0-5872, 1970. 

Cochran, T.H. and Maslca. W.J.. "An Investiga- 
tion of Gravity Effects on Laminar Gas-Jet 
Diffusion Flames," Thirteenth Symposium 
(International on Combustion . The Combustion 
Institute. Pittsburgh, PA, 1971. pp. 821-829. 

Haggard. J.B., Jr. and Cochran, T.H.. "Stable 
Hydrocarbon Diffusion Flames In a Weightless 
Environment," Combustion Science and Technology, 
Vol. 5, 1972. pp. 291-298. 

Cochran, T.H.. "Experimental Investigation of 
Laminar Gas Jet Olffusion Flames 1n Zero 
Gravity," NASA TN 0-6523, 1972. 

5. Haggard. J.B.. Jr. and Cochran. T.H., "Hydrogen 
and Hydrocarbon Diffusion Flames 1n a Weight- 
less Environment," NASA TN D-7165, 1973. 

Edelman, R.B., Fortune, O.F., Weilerstein. G. , 
Cochran, T.H., and Haggard, J.B., Jr. An 
Analytical and Experimental Investigation of 
Gravity Effects Upon Laminar Gas Jet Diffusion 
Flames." Fourteenth Symposium (International) 
on Combustion . The Combustion Institute. 
Pittsburgh. PA. 1973. pp. 399-412. 

7. Edelman, R.B., Fortune. 0. and Weilerstein, G. , 
"Analytical Study of Gravity Effects on Laminar 
Diffusion Flames." NASA CR-120921, 1973. 

Haggard. J.B.. Jr. "Forced and Natural Convec- 
tion in Laminar-Jet Diffusion Flames." NASA TP 
1841. 1981. 

Edelman, R.B. and BahadoM , M.Y.. "Effects of 
Buoyancy on Gas-Jet Oiffusion Flames: Experi- 
ment and Theory," Acta Astronautlca . Vol. 13, 
No. 11/12. 1986. pp. 681-688. 



PM*« 



F1g. 18 (Continued) 



0* ?°°* 



10 



Section 6' 
•GAS-JET DIFFUSION FLAMES UNDER HEDUCED-GRAVITY COND.T.ONS" 



M. Y. Bahadori and R. B. Edelman 



ffjwaa»!aSsSS ss = s = ar 



Houston, Texas April 1988 



A^^^wrt*^*" r Il n m n nWH"^"*"™*"" 1 

l unar Rasas (see Section 14). 



i§ 



LUNAR BASES & SPACE ACTIVITIES 
)N THE 21 s* CENTURY 



April 5 - 7, 1988 

Westin Galleria Hotel 

Houston, Texas 



M. 



GAS-JET DIFFUSION FLAMES UNDER 
REDUCED-GRAVITY CONDITIONS 

Yousef Bahadori and Raymond B. Edelman 



Science Applications International Corporation 
Combustion Science and Advanced Technology Department 
9760 Owensmouth Avenue, Chatsworth, California 91JU 



Paper No. LBS-88-014 



88wg@8&s8ffi^f!®R " 



1 v . ■ r - ■ 



ft*8 



Symposium spon S or«d by NASA, AIAA ; .he 1 j-"" *"J™« »£ ^S'can"" 
Society. 



GAS-JET DIFFUSION FLAMES UNDER REDUCED-GRAVITY CONDITIONS 

M. Yousef Bahadori and Raymond B. Edelman 

rnJ C f Ce f App1icdti0ns International Corporation 
Combustion Science and Advanced Technology Department 
9/60 Owensmouth Avenue, Chatsworth, CA 913H 

buoyancy-inJ^c^Sion'anS t „ -?° uc "e TltTuLU IffecT? ^ bM 
reduced-a levpl* tho fi 3mn , ^ 9 7 J O 'jrusion tiames has shown that under 



INTRODUCTION 



sarety as well as the fundamental understanding Solid limiiri Ln.L 

LAMINAR DIFFUSION FLAMES 

cated. Coupling exists between chemical kinetics, fluid dvnam cs radiation 

the rathe? dfffprpnJ IS ^ V5? remainln 9 Phenomena are responsible for 
the rather different behavior of flames which is observed in micro-gravity ex- 




DIFFUSION 



RADIATION 



FLAME STABILIZATION 



Fig. 1. Physical and chemical processes occurring in laminar gas- 
jet diffusion flames of hydrocarbons. 



2.9 - 



9 


' <■> i 

- i 


Extinction 
\ 


■ 


} ' f 


• 
• 




r ' 


• 


5 


L 

- Fuel flow rate "• 
- 2.8 cc/s [ 


• 
• 




- Nozzle radius i 


• 


1 


- 0.186 ca I 

1* 
i 


• 
• 




!'• 


• 
Ztro-9 


7 


• i u 


- r 1 1 



2.1 - 



4.5 



4.9 5.3 

Time, sec 



5.7 



3.5 



- 2ero-g / 

- length • 



state, f '• - • • 



5 3.0 



8 



_•, 



2.5 



2.0 



':-> 






(b) 



Fuel flow rate 
- 2.1 cc/s 

Nozzle radius 
« 0.051 ca 



— Ave. normal -g 
length 

.Hln. zero-g length 






V-^o 



Time, sec 



Fiq 2. Methane-air flame lengths as a function of time in micro- 
gravity for (a) an extinguished flame [1], and (b) a flame 
which reached an apparent steady state at the end of the 
drop [4]. 



pen'ments conducted to-date Tn thi « t^HAn tu„ u - 

(Moon) conditions are presented. 9 l^ars; ana l/6g 

drop ^r , ?.2i[, , ;; , ?9"r."?" S .«« , l ^gTt^c^ d' 2 " ^"^ 
Mrocarbon flames we?e ionfted Ind «t»h i.il fl hydrogen and various 

« STtK rr^.r.I'S dro/?^ T™ »^' -n^arl^ad/ ?.M 

UWK ss&^yj'- "« *$ ^ * ^.«™ "»" 

transitory state bvt^'tf™,; 9 a V Ppi,re,,t ste "y state, or remaining in a 

A mathematical model has been developed T6 91 fnr the cf„Hw «* i 

inlrtl J coI?tv H.fi 6 e < uatl0 " s of mot1 'o". which includes the effects of 
hows ?he exrp? p^; J lffusi <>n. radiation, and chemical reactions. Figure 5 
under both nn™? a S r f™nt between the predicted and measured flame heights 
fl It "O^l-gravity and zero-gravity conditions [6, 91 We have recent?* 

' loure 6 shls lu^ °l mt ^ M V™* Under ^-en^itl * 
distance rv !J? n ?"-f ^nsional centerline velocity vs. the axial 
1 ■ Clearly the effect of convection plays a major and different role 
for different gravitational environments. The partial effect of convection 

oT" on and rLi a n t? 0rmal - 9 'ir dC0Upled Wlth ?he «"""«' ^ " ot 
arav a t^?on a d ™ dlatlon ' result * ™ different flame behaviors under different 

norml a a rav a iv ,nS mi ^ 6Xa ? Ple ' Pig - 7 ShowS a P ro P ane flame "^ both 
nrnJurt? JJ y and ™C"o-gravi ty conditions. The shape, color, luminosity 

SnSe tnn2 „n?f USt10n * and in genera1 « the flame «*"!» or canno tbl ?ru y' 
Ztl l» U ? l6S ^ e *P eriments ^e conducted (along with appropriate diagnostics) 
under the g-level of interest, and then when combined with the theoreUcal ' 



20 



16 



12 



s 



— Transient flaaes («> height available) 
...No data available 
Data available 



Data[4] J ± steady-state flaaes 
• Extinguished f lanes 



D Flaaes Ignited 1n 
pg and reached 
steady state [8] 



* 4 



r - 0.0825 c« 




20 



40 



60 



80 100 120 
Reynolds Nunber 



140 



160 



180 



200 



Fig. 3. Methane flame heights as a function of fuel Reynolds number [8], 




12k 

11 

10 

a 



"1 * 7 






• • • • * 

o o ° ° ° ° o 



■(b) 



o o 



* * 



• o 



*• . . . B S 3 8 5 8 B 8 

■ n O 




u**l 


IM1»<*) 




• 

■ 
O 
A 

o 


7 

t.mn 

I 


I.t 
I.* 
• » 
1.1 
1.0 
» 



TVm (mc) 



i.a i.e a.o 

Tbna (nth 



2.4 J.8 3.2 3.6 



Fig. 4. 



Micro-qravity flames of (a) methane and (b) propane [8]. The low-volume 
now rate flames of methane did not extinguish, but were extremely faint 



and difficult to observe. 



1000 r- 



100 - 






10 - 




100 



Re„ 



Eip«rIm«nroi doto 



1.4 



Nor mol - g 


r , cm 


O 


0.051 


A 


0825 


D 


113 





186 


o 


318 


7 


44? 



Solid symbol* denott 
2910- g 



J 



1000 



Fig. 5. Comparisons between the theoretical predictions and experimental 
Sane-a" fUmes^' " 1 fU ™ h *^ vs ' Jet Reynolds n Umbe 



rs- 




Fig. 6. Centerline velocity vs. axial distance along the jet for a methane- 
n J>\ " ? fUnCtl0n 0f g rJvi ^tional le?el. Nozel radius = 
46 8 cm/s' fl ° W rate = ' CC/SeC " preSSUre = l * tm « "o « 



analyses, will provide a data base for the understanding of fires both on earth 
and under reduced-gravity conditions such as on the Moon, Mars, or in free space, 
"a recent publication [10], we have analyzed the past experimental and theore- 
ticJl efforts on laminar gas-jet diffusion flames. The following section gives 
a brief account of the potential for conducting turbulent diffusion flames under 
reduced-gravity conditions. 

TURBULENT DIFFUSION FLAMES 

Turbulent gas-jet diffusion flames under normal-gravity conditions have been 
the subject of extensive theoretical and experimental studies for a number of 
decades. The classical behavior of a turbulent jet diffusion flame is shown 
in Fiq 8 for the dependence of length and structure of the flame on flow vel- 
ocity for a tube of given size. As the jet velocity increases, the flame tran- 
sists from laminar (where mixing is governed by molecular diffusion only) to 
fully developed turbulent behavior (where mixing is due, largely, to eddy dif- 
fusion or convection, with the final homogeneity being attained by molecular 
diffusion). For the tube size used in the flame study of Fig. 8, a velocity is 
reached where further increases in the jet velocity result in no change in flame 
heiqht. It can be shown that this is a characteristic of momentum-dominated 
flames, i.e., flames in which buoyancy is not important. Although much progress 
has been made toward the characterization of momentum-dominated turbulent flames, 
i.e., flames with large Froude number (Fr = v2/gd, where V is the flow velocity, 
q is the gravitational acceleration, and d is the nozzle diameter), this is not 
the case for low-momentum turbulent flames characteristic of unconfined fires 
In this case, the fire research community depends primarily on empirical results 
which, having been obtained under normal-gravity conditions, have the buoyancy 
effect inherently embedded within these correlations. When buoyancy is impor- 
tant (i.e., low-momentum flames), unlike Fig. 8, a constant height as a function 
of velocity is not reached for fully developed turbulent region (see Fig. 9). 
Although there is an apparent dependency on Reynolds number (Re = VdA >> where V 
is the flow velocity, d is the nozzle diameter, and v is the kinematic viscosity) 
beyond the transition region, this is really not the case The reason is that 
as the nozzle size gets sufficiently large for a given velocity level .the buoy- 
ant force becomes important; this is reflected through the J™«« ""^ l; e -! 
the ratio of momentum to buoyant force. In fact, normalization of t^^scribing 
equations shows that in general, for fully developed turbulent flows the flow 
should be independent of Reynolds number but not the Froude number Although it 
appears that the buoyancy effects are indeed important as the momentum level de- 
creases relative to the buoyancy effect (i.e., « Froude number decrease ), the 
mechanisms responsible for this behavior are far from being fully ^erstood. 
Thus, the need for more fundamental data and analysis is apparent because or 
the requirement to define the hazard and control of fires not only on earth but 
under reduced-gravity (e.g., Lunar and Martian bases) and micro-gravity (i.e., 
in space) conditions as well. 

For low-momentum flames, strong interactions between buoyancy and J;^ulent- 

flame structure exist which affect the flame behavior and chemistry. There are 

wo gr J- ndu ed mechanisms responsible for the alteration of a low-momer ituj. 

flowfield The first arises directly from the buoyant force acting on the time- 

eraged or £aV flowfield and appears as a gravity term in the mean momentum 

equation. The second mechanism arises out of the int % a ^^"^J we k f ne Jtc enerqy 
and velocity fluctuations which manifest in a source of turbulent kinetic energy 
Snder normal -gravity conditions, it is not possible to separate these two effects 
in terms of their impact on mixing rate, and hence, flame structure. Clearly, 



lull red, 



Yellov 



Yellow 



•Cherry red 
Orange 



Bright J 




1 / \////\ 


White ~\ 


V 'rf/A 
\/ {///A 




White with 






more yel low 




(a) If 


Mixed blue,/ 
pink, orange, 






yellow toward 


center 



-White with little yellow 
•Violet/pink 



Fig. 7. 



il\ H l[7 ] Tn a 7 lil y a U d (b) ; icr °-g«vity propane-air diffusion flanes 

aL L r3dlUS = °-° 825 Cm and volume f] ™ rate = o S ec 

bcale. i — i = 1 cm; reproduced from [81. ' 



Laminar 
flames 



Transition 




Fully developed 
turbulent flames 




/^Envelope 
Oof < 
K flame \ 

Y-l?--?~1?- 




Increasing noule velocity 



Fig. 8. Change in the flame height and behavior with increase in nozzle velocity 
for a typical gas-jet diffusion flame [11]. y 




200 400 600 800 1000 1200 1400 1600 1800 2000 

Volume flow rate, cm 3 /s 



Fig. 9. Effects of volume flow rate and tube diameter on flame 
height for city gas diffusion flames [12]. 



l.o 



_« 






OATA: 




L- PREDICTIONS: 

ZERO-6 

BUOYANCY-flEAN FLOW INTERACTION 

BUOYANCY-MEAN FLO* - TURBULENCE 



O.Ol! 



J I I I I I l 



J L 



10 



SO 



X/0 



Fig. 10. Effect of modeling assumptions on the prediction of center- 
line temperature for a hot buoyant jet (U. = 21.6 ft/sec); 
data from [13]. J 



the advantage of operating in a low-gravity environment would be to provide a 
major significant base of new information isolating the combined effects of 
buoyancy. Then, with a theoretical model, the effects of buoyancy on the mean 
flow and on the generation of turbulent kinetic energy would be separable. With 
the aid of a theoretical model characterizing these two mechanisms, recent stud- 
ies at SAIC have illustrated the importance of gaining a further understanding 
of both of these mechanisms. The model has been applied to a hot plume in which 
temperature measurements were made within the plume. The results are shown in 
Figs. 10-12. Figure 10 shows a comparison between the relative effects of the 
two buoyancy-related mechanisms (i.e., effects on the mean flow and on the gen- 
eration of turbulent kinetic energy). The data points correspond to the normal - 
g condition [13]. The best agreement with the data is obtained when both mean 
flow and turbulent interactions with buoyancy are taken into account. Figure 
11 shows the non-dimensional centerline distance as a function of Froude number 
for a particular value of non-dimensional centerline temperature. Comparison 
with data is again satisfactory, and the effect of Froude number on both normal- 
g and zero-g (Fr = ») cases is shown. Figure 12 shows the effect of Reynolds 
number for both normal-g and zero-g conditions. Comparisons with the data show 
good agreement and as the Reynolds number increases, the behavior approaches 
that of the zero-g case. (The data points close to the jet port show a poor 
agreement with the predictions which is believed to be due to the disturbances 
caused by the heating coil mounted inside the jet set up). As Figs. 10-12 show, 
the effects of buoyancy on both turbulent kinetic energy and on the mean flow- 
field are equally important. Numerical models would be capable of separating 
the effects of the two mechanisms, and when combined with the experimental re- 
sults under both normal-g and arbitrary-g conditions, would elucidate the be- 
havior of turbulent diffusion flames under a variety of reduced-gravity condi- 
tions such as Martian conditions (2/5g), Lunar bases (l/6g), and in space 
(<10" 5 g). 

The results presented in Figs. 10-12 are for an externally heated plume. 
What is required now is application of similar turbulent models to flames 
wherein we might expect the effects to be even greater than those in the hot 
plume because the distributed heat release in a flame results in the persis- 
tence of gravity-induced mean flow and turbulence under normal-g conditions. 
Thus, we would expect the difference between normal-g and reduced-g environ- 
ments to produce noticeable and significant differences in the flame struc- 
ture. Of course, an experimental data base under reduced-gravity conditions 
would be invaluable in furthering our understanding of these effects on the 
one hand, and the question of fire behavior and prevention on the other, in 
addition to the development and validation of the theoretical models account- 
ing for the two mechanisms previously cited. 

CLOSURE 

Although the effects of buoyancy on low-momentum flames have been qualita- 
tively observed, it is only recently that quantitative descriptions of the 
phenomena affected by gravity have been attempted. Understanding this pheno- 
mena is not only of fundamental interest but it is of critical importance to 
fire safety in space as well as on earth. Furthermore, for processing and 
manufacturing in space, controlled flames used as one form of heat source are 
likely to be employed. This paper has presented results that indicate sign- 
ificant effects of gravity on laminar and turbulent flame structures. More- 



_• |o 



Fig. 11. 




Non-dimensional center! ine distance as a function of 
Froude number for a particular value of non-dimensional 
centerline temperature in a hot buoyant jet (U^ - 21.6 ft/secj, 
data from [13]. 



1.0 



-aagr 



-» 



0.1 



-» 



0.01 






[-t KSU.TS 




MTA 
SntOL 


PDEDICTION 
CUM 


*», 









MM 

VK 
IOC 





0-C KSULTS 







Ml 



I 'till 



X/l 



10 



J L 



SO 



Fiq 12. Effect of modeling assumptions on the prediction of «nter- 
9 ' line temperature under both normal-g and zero-g conditions 
for a hot buoyant jet; data from [13]. 



10 



1, 

2. 



4 
5. 
6. 



7 
8. 



9. 
10. 



data obtained under reduce^-JraJ ^ quantitative 

sue limitations is needed. 9 T h poten?? ? ] \iV - hl i Ued by test time ™* 
conducted on the Moon and Mars offers th?l n . " d ° ta from "Pertinents 
equalled in earthbound facHitie inclu no T^ 1 ^' 0ne whlch «nnot be 
parabolic trajectories. C1Mtles 1ncludl "9 drop towers and aircrafts flying 

Acknowledgemp ntc; . Thl - S , . . 

pteTliHSir-lhe- pac ^ Phjs c and K-'; 'V^ ^NASA-Lewis Research 
Contract NAS3-22822, w t M Sa ndr Sn^T*"* 5 1n S P ace ) Program, 
the NASA project scientists ° 1S ° n and Mr ' Dennis P- Stocke? as 

REFERENCES 
Cochran, T.H. and Masica W .1 "rf foftr nf r 
Diffusion Flames," NASA f N 0-5872^1970 Uy 0n Lam1nar Gas " Jet 

Cochran, T.H. and Masica w i »a„ T „ *• .. 

Laminar Gas-Jet Diffusion Hames J \J? T t, ?"' on of G "» ( 'y Effects on 

^o^stio^ The Co^s^on^sti^MM^ 

PP. 291-298, 1972. scien ce a nd Technology . y J. 5, 

natrin^o ^^ n SL , reSr!9?J. L " U " r GSS - Jet ° iff — 

"^in^^u^ L°^n^ 

Edelman, R.B., Fortune, F Weilpr<t -in r r u 

"S* NASA'TpriMuMsl. 3 "' N " Ural C ° nVeCt '°" '» Laminar-Jet Diffusion 
Oif^Jon'F^^tde^i^Gr^nJ'coniu, 3 ^ **?*'> °- P - ^»'»" 

"?ic? s \„ R -? am in^-??; u ?: r^-i::^-c R G :i^r,^^: stu ^ ° f —> 

Edelman, R.B. and Bahadori. m y «cff flr tc «* d 



ii 



11 Hottel, H.C. and Hawthorne, W.R.. "Diffusion in Laminar Flame Jets," 
U - ^ sym posium on Combustion , pp. 254-266, Williams and Wilkins Co., 
Baltimore, 1949. 
„ „ li v r^w r and KaDD N , "Diffusion n^." Third Symposium 
12 - rkmLt^l li C 288-300: P Silliams and Wilkins Co., Baltimore, 1949. 

13 Cleeves, V. and Boelter, L.M.K., "Isothermal and Nonlsothermal Air-Jet 
Ir^^Hnn,." Chemical Engine ering Progress, Vol. 43, No. 3, pp. 123 

133, 1947. 



12 



SfifillflHl 



•OXYGEN-CONCENTRATION EFFECTS ON MICROGRAVITY LAMINAR 
METHANE AND PROPANE DIFFUSION FLAMES" 



M. Y. Bahadorl and D. P. Stocker 



Paper presented at the Eastern States Meeting of The Combustion Institute, 
Paper P resemTO A ° bany New Yo rk, October/November 1989 



Oxygen-Concentration Effects on Microgravity Laminar Methane 
and Propane Diffusion Flames 

M. YousefBahadorit 
Science Applications International Corporation. Torrance, California 

and 

Dennis P. Stocker 
NASA-Lewis Research Center, Cleveland, Ohio 

] Introduction 

Laminar diffusion flames of hydrocarbons under microgravity conditions have shown distinct characteristics 
Sfto normal-gravity flimes [1]. Compared to the flames in normal-gravity environments larger 
sootier and somewhat globular flames are observed in microgravity. This is due to the significant reducoon 
SS bSStoS wWch makes diffusion the dominant mechanism of transport As a result increased 
resfdenS toe, Snced soot formation, radiation cooling due to the larger flame size, and the possible 
Set ofa chemical-kinetics limitation on the heat-release process become responsible for the very different 
characteristics of these flames compared to those in normal gravity. 

In normal-gravity environments, laminar diffusion flames of hydrocarbons flicker and are yellow. ^whereas 
£e™ microgravity counterparts are orange/red with practically no flicker. The colors curved in the 
obvious Ss [ 1 ] indicate that extensive soot formation enhanced by the increased residence tune result 
?r ?mon SSS and hence, delayed soot bum-off downstream of the flame In addition, open tips are 
obs^ld^r propSe K?s [21. which indicates that unbumed and pyrolyzed fuel spec.es escape through 
u,e flame tip^n microgravity; Ihis tip-quenching combined with thermophores.s are responsible for the 
observed underventilated-type microgravity flames. 

The previous studies [1. 2] have been limited to atmospheric flames burning in air. In this paper, we 
Sesem mr r Lu?ts of m crogravity experiments conducted for methane and propane diffusion flames wuh 
^ygen-^rogen mixtures containing 18% - 30% oxygen, and burning in quiescent environments at 1 atm 
pressure. 

? Apparatus and Experimental Procedure 

The experiments were conducted in the 2.2-second NAS A-Lewis Research Center Drop Tower. Schematics 
of Spararus and a detailed description of the experimental procedure can be found elsewhere [2^Two 
nleTnozdS were used in the studies.^ tapered tips had a nozzle exit radms of, W™™^ £048 cm 
and f) 074 cm resDectively The flames were ignited in microgravity using a spark electrode located at 1 .0 
^ abSve S*S tip and 1 5 cm away from the nozzle centerline. with the nozzle acung as the ^second 
etectrode The spark ignitor, unlike the one in the previous study [2], was far enough from the flame 
S^to^SSflSe disturbance. As .result, a stronger ^^I^^S^SSZ £ 
was used to ensure that low-oxygen concentration flames indeed ignite. ThejP^; , « mUon s y stem was sel 
to 0.4 sec. to 1.0 sec. depending on the flame, following the release of the package. 

Methane and propane diffusion flames were studied at 1 atm chamber I""" «** °2^2 ■£*»*»£ 
i so, 19<& 21% and 30% oxvaen The volume-flow rates were 1.75 and 5.25 cc/s for methane, and u.m 
Ld 6 96^c/s Spropa^ ^fiEning rate was 12-16 frames/sec., and the CW^gS^ 
Video News) films were force -processll 2-f stops for low-oxygen-concentrauon and 21%-02 flames. The 
slow filming rate and force-processing ensure that blue is discemable. 



* Presented at the 1989 Fall Technical Meeting (Eastern Section) of the Combustion Institute. Albany. 
New York, October 30 - November 1, 1989. „- 

This wot is supported by NASA-Lewis Research Censer under Contract NAS3-22822. 

t Author to whom correspondence should be sent; on-site at the NASA-Lewis Research Center. MS 500-217. 
Cleveland. Ohio 44135. 



3. Results and Discussing 



Figures 1 and 2 show the name height as a function of percent oxygen for differrnr flnu, «,,.. a w t 
normal-gravity and microgravity conditions. The methL rcidu CRg ,) £ fo ftS'n'f 
The propane results (Fig. 2) correspond to both small and large nozzle? X flaSa^w X ' 

Drastic changes in flame color and luminosity occur due to the absence of buovanrv Thic „„k. r 

SSES^T l ? nai £ b, y m toth "tor and characteristics. Referring to Fig. 3?tf>"dim E£b£ flS 

S °° k niore-or-less the same from the color point of view. The trends for pmm^^SXlt 
c^S^SrylS 

The unique underventilated behavior of microgravity propane flames shows that extensive soot formation 

fn^/r 1 /, 5 ' 5, "? **? residcncc dinc «* ««■*» cwtai i« > characSc of S^virv^L 
£ ?n^ C fiIms K !? ow ? at f 001 "capes through the tip of the flame (see red and dTSKS) TOs 

Mc^£^SS£ ^ f Umited oxygen ; fu ' 1 mbdn * «»*** to«. ICf iff i3Sbu5S; 

ff" c ,l™ soot Parades, and extensive amount of unbumed fuel escaping thorugh the flame ti D The 

HwZFtetSZTJ™' ° f W™*™ *« »«-»i"» citufctionTay ri wrSfch^g 
However, the extinction mechanism may be different compared to normal-gravity flames- in other word7ar 

S^T^ Xgen . concentrat i ons we ™y observe a flametfthe same siK soSo "a^K Tafhea 
release due to reacoon can no longer compete with heat loss, causing flame extinction^ 



second Zero-Gravity Facility of NASA-Lewis. 
aq Qidusions 

cause of flame extinction for the near-limit flames. 



References 
1. 



2. 



sssta asraiSiS^>tar?M: saws k 

references cited in this work. 

Bahadori. M.Y.. Edelman. R.B.. Stocker. D.P.. and Oton^SU "IgjAon and Behavior of 
Laminar Gas-Jet Diffusion Flames in Microgravity." in press. A1AAX. 1989. 




16 




14 


- 






U 





10 



1 » 




_1 I l_ 



■ » I 1 L. 



16 



U 



20 



22 



26 



26 



28 



12 



Fig. 1. Flame height vs. percent oxygen 
for both normal-gravity and 
microgravity methane diffusion 
flames. 



Fig. 2. Flame height vs. percent oxygen 
fa both normal-gravity and 
microgravity propane diffusion 
flames. 



Methane Flame 
191 Oxygen 
1 atm. 

Small Nozzle 
5.25 cc/s 



Methane Flame 
211 Oxygen 
1 atm. 

Small Nozzle 
5.25 cc/s 




Methane Flaroe 
301 Oxygen 
1 atn. 

Small Nozzle 
5.25 cc/$ 




0-g 



0/R 



Propane Flame 
181 Oxygen 
1 atm. 
Large Nozzle 
0.96 cc/s 

/ I 



Propane Flame 
211 Oxygen 
1 atm. 
Large Nozzle 
0.96 cc/s 



OR 



1-9 




Propane Flame 
301 Oxygen 
1 at*. 

Small Nozzle 
0.96 cc/s 




Fig . 3. Normal-gravity and microgravity laminar diffusion flames of methane and propane, 
burning in quiescent environments at 1 atm with different oxygen-nitrogen 
mixtures. The two nozzles have a tapered-tip radius of 0.048 and 0.074 cm. The 
various colors indicated in the diagram are as follows: B (blue), D (dark). O 
(orange), P (pink), R (red), W (white), Y (yellow), and DR (dull red). The bars 
show the range of normal-gravity flame flicker (0. The coloring may be slightly 
off due to the force-processing of films for low-oxygen and 2 1 %-C>2 environments. 
The films for the flames in 30%-O2 environments were not force-processed. The 
low-oxygen-concentration flame of propane in normal gravity shows pockets of 
flame leaving the flickering part. Scale: i 1 = 1 cm. 



§££tifiIL& 

"MEASUREMENT OF TEMPERATURE IN MICROGRAVITY LAMINAR 

DIFFUSION FLAMES" 



M 



Y. Bahadorl, R. B. Edelman, R. G. Sotos, and D. P. Stocker 



Paper V*^***™*" Ktfa 6 ^ 



Institute, Orlando, Florida, December 



Measurement of Temperature in Microarav itv Laminar Diffusion Flames 

M.Y. Bahadori 2 and R.B. Edelman 3 
Science Applications International Corporation, Torrance, California 

and 

R.G. Sotos and DP. Stocker 

NASA Lewis Research Center, Cleveland, Ohio 



1. Introduction 

Laminar diffusion flames of hydrocarbons under microgravity conditions have been studied extensively [1-11]. It has 
been shown that compared to the flames in normal-gravity environments, larger, globular and more sooty flames are 
observed in mcrogravity. This is due to the reduction of buoyancy, which makes diffusion a more doming ,nt 
mechanism of transport. As a result, increased resklence time, enhanced soot formation, radiatrve cooling due to the 
larger flame size, and the possible onset of a chemical kinetcs limrtatwn on the heat release process become 
responsible for the different characteristics observed between normal-gravity and microgravity flames. 

In the Drevious studies [1-11], qualitative information and semiquantitative data have been obtained using visualization 

Center. However, the limitation of available time in microgravity and the lack of quantitative data have prompted the 
current investigation. 

In the present work, laminar diffusion flames of propane burning in quiescent oxidizing , environments ; ol f different 
pressure and oxygen concentrate have been studied in the 5.18-Second Zero Giavfty Fac.lrty of NASA Lews 
Research Center The first demonstration of the ability to map the thermal field in a microgravrty combustion study 
is presented in this paper. A rake of nine thermocouples has provided temperature data. Cinematography atong with 
this data has enabled us to compare the flames quantitative^ and obtain information on their characteristics. For each 
microgravity flame, the normal-gravity counterpart was also studied. 

2. Apparatus and Experimental Procedure 

Figure 1 shows a schematic of the hardware. The volume of the sealed combustion chamber is 0.087 m 3 . The tapered- 
tip nozzle inner radius is 0.0825 cm. Propane flames at a volume-ftow rate of 15 ocftec : were st udied. The coW-jet 
Reynolds number was 1 30 in all cases. The oxidizing environment was composed of 1 9%-<V81 %-N 2 at 1 .0 atm, 21 /«- 
0/79%-N 2 at 1.0 atm, and 21%-02/79%-N 2 at 0.5 atm, respectively. 

The flames were kjnrted using a spark electrode, located 1 .0 cm above the nozzle tip and ™™W<^™»* 
centerline, wrth the nozzle acting as the ground electrode. The spark-.gmt.on system was *«^°2ZTr7>*l 
1 sec for both the normal-gravity and microgravity flames. The microgravrry flames were gnrted after the release 



1 Presented at the 1990 Fall Technical Meeting (Eastern Section) of The Combustion Institute, Orlando, 
Florida, December 3-5. 1 990. This work is supported by NASA Lewis Research Center under Contract NAS3- 
22822. 

2 Author to whom correspondence should be sent to the following address: SAIC/NASA LeRC, MS 500-21 7, 
Cleveland, Ohio 44135. 

3 Present Address: Rockwell International Corporation, Rocketdyne Division, Canoga Park, California 



of the package in the evacuated drop tube. 

Two movie cameras (filming rates = 12 and 30 frames/sec, respectively) using Kodak Ektachrome Video News films 
recorded the flame development and behavior. A rake of nine thermocouples (3x3) measured the temperature above 
the normal-gravity flame tip and both inside and outside the microgravity flames. These thermocouples were located 
m a plane passing through the nozzle centeriine. The thermocouple columns were 2 cm apart with the column closest 
to the nozzle centeriine being 0.75 cm away from it. The rows of thermocouples were 3 cm apart Although the 
thermocouples were fixed with respect to each other, the whole rake could be adjusted vertically with respect to the 
nozzle tip (prior to testing) to measure temperature at specified heights. 

3. Resurts and Discussion 

Figure 2 shows the size, shape, and color of the flames studied in addition to the position of the thermocouple rake 
with respect to the flame. The normal-gravity flames were typically established within 0.5 sec following ignition and 
fleered throughout the experiment duration with the typical flicker frequency of laminar diffusion flames burning in 
qu«escent environments. All of the normal-gravity flames were generally yellow/white, and the color was not stronqlv 
affected by either pressure or oxygen concentration. The results have shown that the lower the oxygen concentration 
or pressure, the wider the normal-gravity flame. This can be explained in terms of the availability of oxygen in the 
environment. The results presented in Fig. 2 also agree with previous observations that at a fixed Reynolds number 
the lower the pressure, the smaller the flicker range [9], and the lower the oxygen concentration, the larger the flicker 
range [7] for normal-gravity propane flames. 

Flames in microgravity were always taller and wider than their normal-gravity counterparts, and were flicker-free 
Removal of buoyancy results in the dominance of diffusive processes which are responsible for these changes In 
addition, the lack of buoyant force causes the flame to develop toward its steady state over a much longer time (of 
the order of seconds). The microgravity flames shown in Fig. 2 correspond to the shapes and colors observed just 
prior to the deceleration of the package, and were fairly uniform during the last second of the drop. However unlike 
flames in normal gravity which were pencil-like and had closed tip, the microgravity flames of propane had an open 
tip, resembling underventilated behavior. This is believed to be caused by the combination of enhanced sooting a 
significant drop in the gas velocity near the flame tip, extensive radiative loss, accumulation of the combustion products 
in the vicinity of the flame, and thermophoretic effects. These processes cause the flame tip to open due to the 
reduced transport of oxygen, and presumably result in fuel escape through the tip. The microgravity flames of Fig 
2 are dimmer, more sooty, and orange/red in color. Just like normal-gravity flames, the lower the oxygen concentration 
or pressure, the wider the flame. Using the criteria of red-to-dull red transition region for the location where burning 
ceases [9], it can be seen that the lower the oxygen concentration or pressure, the taller the microgravity flame [7,9], 

Figure 3 shows a typical set of temperature data for the propane flame burning in 19%-0/81%-N. environment at 1 
atm, under both normal-gravity and microgravity conditions. The lowest row of thermocouples is 10 cm above the nozzle 
tip. An initial overshoot in temperature is observed in both cases due to the presence of excess fuel at ignition. 

Figure 3 presents some useful information not only in relation to the temperature field, but also for other flame 
characteristics and behavior. The normal-gravity data show that the temperature drops quickly to the ambient 
somewhere between the radial locations of 0.75 and 2.75 cm. This is due to the effect of buoyant force which carries 
and removes the combustion products in a narrow column determined by the maximum flame radius In addition the 
strong buoyancy-driven portion of the flame causes the three thermocouples near the centeriine to show roughly equal 
temperatures, which essentially reach steady state approximately 3 seconds after the start of the experiment. 

The data presented in Fig. 3 for microgravfty flames show that far above and away from the flame, the gas is still 
expenencing a temperature rise due to the continuous release of the combustion products in the vicinity of the flame 
and their subsequent diffusion. This figure also shows that the gas temperature does not reach a steady value 



anywhere in the field during the 5 seconds of microgravity, again due to the continuous dilution and heating of the 
SmeS caused by the combustion products. In addrtion, large temperature gradients exist m the axia direct on 
or e m a gravrty flame, and the variation of temperature in the radial direction is secant all the way to the far 



field. 



Fiaure 4 shows the temperature data at t - 5 seconds as a function of the axial locate of the thermocouples for fixed 
S loc^ on Bo h n'ormal^ravity and mtorogravrty data are shown in this figure All of the ~^™*d Ub 
nd cSt the second and third columns of thermocouples (see Fig. 2) reg.ster a temperature of 300 K. The is 
Smn (closest to the centerline) shows a constant and small decrease in temperature wrth an increase ,n height : fw 
Ses A B and C Normal-gravity flame D does not show a fixed slope, probabry because rt is a taller, wider, and 
mo™ dfluse ame dVe to the tower pressure. The mcrogravrty flames A and B of Fig. 4 show almost .dent** 
Zei; SLtons but rt needs to be determined whether flames in environments wrth less than 1 9% oxygen 
wTshow h ^ same levels of temperature. Significant dHerences in temperature levels are observed between 
mloXity flames C and D, wrth the 0.5-atm fiame being taller and wider. The combinatton o. larger flame s.ze and 
TeSVessure contribute to increased radiatrve toss, which raises the field temperature to hgher levels. 

Fiaure 5 shows the temperature data as a function of the radial location of the thermocouples for fixed axial locations 
T mi og^ ^ fames. The normal-gravrty data is not presented in this figure, since the temperature , drops ; to 
ambiert somewhere between the first two columns of thermocouples. Due to the uncertainty in the monoton.c behav,or 
o\1em^« bileen the first and second thermocouples of the towest row, the plots i havenot been extended to 
tta i temperature at the first thermocouple. It is possible that there is a peak in temperature between the mnermost 
thermocouples due to the presence of the flame zone, especially for flames C and D. 

4. Conclusions 

Mapping of the thermal field has been demonstrated for the first time in microgravrty flames using a rake of nine 
thermocouples. 

Normal-gravrty and microgravity flames of propane burning under different environments of pressure and oxygen 
concentration show significant drtferences in height, width, cotor, and temperature d«stnbut.on. 

Low-pressure microgravrty flames show significantly higher temperatures; this is apparently caused by both increased 
radiative loss and larger flame size due to reduced pressure. 

Compared to mtorogravity flames under atmospheric condrtions. the low^xygen-concentration flames do not show as 
tigXrtan effect as the low-pressure flames. However, these oxygen-deficient flames may show signrfcant changes 
at low-enough oxygen concentrations. 

Longer test durations appear to be required to reach steady state in microgravrty. 



References 



1. 
2. 
3. 
4. 



5. 
6. 

7. 



8. 

9. 
10. 
11. 



Cochran, T.H. and Masica, W.J., "Effects of Gravity on Laminar Gas Jet DWusk>n Flames," NASA TN D-5872 
D^^ngTl'i ^P 6 "™ 13 ' lnvesti 9*on of Laminar Gas Jet Diffusion Flames in Zero Gravity," NASA TN 
^^^^^ ^ ~™ D ^ »™ ■" - Weightless 

mmfn&oiL Combustion, pp. 399-412, The Combuston ^J^^' p^^T^^^ 
Haggard, J.B., Forced and Natural Convection in Laminar Jet Diffusion Flames " NASA TP imi mqai \ 

Theory, Acta Astronautica. _£ No. 1 1/12, pp. 681-688 (1986) H«"'"eni ana 

pfn^n^: 'n St ° Cke ^ D,P - "^en-Concentration Effects on Mcrogravity Laminar Methane and 

sssss; %& n v* n and Behav " of Laminar Ga " et 

Steta DP'^nnS? * A B er ! SP f e SaenCeS Meetin 9' Reno ' Nevada - Jan ^ 1990- 
btocke D.P., Size and Shape of Laminar Burke-Schumann Diffusion Flames in Microqravitv " Sonno 1990 
Central States Meeting of The Combust™ Institute, Cincinnati, Ohio May 1990 ^^ ^ m 1990 
Bahadon, MY., "An Analytical Solution for Transient, Cylindrically Symmetric Uminar Diffusion Flames in the 
Z7m BU ° yanCy; Sprin9 1 "° Centra ' St3teS Meeti "9 of ^ Huston 42 ^ Shfo 



THt«K3C0J>.£S 



S"JU (LICTMN 




>*«« UICTKJDC 



mm cMtu 




mitt unm 



futi mini 



Figure 1 . 



Experimental hardware for the 5.18-second tests. 




Figure 2. 



Flames of propane at 1 .5 cm 3 /sec for both normal-gravity -ml ^mjcrogravjty ^conditions ■ JJ J*J™ 
« f«iio>«- Rmiu^ D (dariO DB (dark blue), OR (dull red . O orange). P (pink), R (red), w (wnite), 
Y ( eS' ^eS^r^'f, is shown for the norma.-gravity flames. L is the distancebetweer , the 
nJSeTpan^the Slest £/oi thermocouples. The co^mns of tnerrnocouptes are -spectiveiy 07 
cmaway from the centerline, and then 2 cm apart. The rows of thermocouples are 3 cm apart. These 
diagrams are to scale. 



0.75 aa I 



NOT TO 
SCALE 



.2 cb ■! cm 
i» ■ ■ i 



T 



■t-tv 



f J 



3 cm 
3 cm 



L - 10 



Propane -19S 0-, Flame 
Flow Rate -1.5 cc/s 

Noizle Radius ■ 0.0825 m 

f ■ lita 



Normal Gravity 



Micrpgravi ty 



1200 



200 



~^^=- 



t (sec) 



EE 



- - w - 



1200 



200 



t (sec) 



^ 


^ 


H 






^ 

-~~1 



1200 












1 












































>- 






















u f* 






















.i" 4 








e.f 




200 










C 















t (sec) 



1290 



200 



? 



t (sec) 



ESI 



TTV; 



I 1 



1200 



200 



1 












■ 1 








— 2— . 

h.i 




■ 



t (sec) 



1200 



200 



t (sec) 



-1_L 



§ 



Figure 3. Typical temperature data for a normal-gravity and microgravity flame. 



1200 
1100 
1000 

900 
800 
700 

600 
500 
400 
300 



p = 0.5 atm 

211 2 
L > 8 cm 



Figure 4. 




t\on 



:\rrr 



R • 4.75 era 



■ I 1. 



8 11 1« 




11 14 10 

Axial Location (cm) 



Temoerature olots as a function of the axial location of thermocouples for both normal-gravity (open 
I;Z,s a" Splines) and microgravity (cfcsed symbols and ^") ^ £3^ * 
a volume flow rate of 1.5 cm 3 /sec. L is the distance between the nozzle 1 p^ and the lowest row 
thermocouples. The radial locations Indicated in the d.agram are apphcable to all plots. 




Axial Location 
J >8ci 



-*— « "- 



P » 1 atm 

211 2 
L « 8 cm 


• B 








P * 1 atm 
191 2 

L ■ 10 cm 



- A 



Axial Location 
X » 10 cm 




,1 L. 



■4— A 



0.75 2.75 «-75 



0.75 2.75 4.75 0.75 2.75 4.75 
Radial Location (cm) 



P ■ 1 atm 
2 IS 2 

L ■ 10 cm 




0.75 2.75 4.75 



columns of thermocouples. 



Section 9* 

•IGNITION AND BEHAVIOR OF LAMINAR GAS-JET DIFFUSION 
FLAMES IN MICROGRAVmr 



M. Y. Bahadorl, R. B. Edelman, D. P. Stocker, and S. L. Olson 



AIAA Journal. Vol. 28, No. 2, pp. 236-244, 1990 



• This paper is a revised journal publication of paper AIAA-88-0645 (see Section 5). 



Ignition and Behavior of Laminar 
Gas-Jet Diffusion Flames in Microgravity 

M. Y. Bahadori, R. B. Edelman, D. P. Stocker, 
S. L. Olson 



* 



Reprinted from 

AIM Journal 

Volume 28. Number 2. February 1990, Pa 9J« 236-244 

AMERICAN INSTITUTE OF AERONAUTICS AND ASTRONAUTICS. INC. 

370 LENFANT PROMENADE. SW • WASHINGTON. DC 20024 



Zib 



AIAA JOURNAL 



VOL. 28. NO. 2 



Ignition and Behavior of Laminar Gas-Jet Diffusion 

Flames in Microgravity 

M. Yousef Bahadori* and Raymond B. Edelmant 
Science Applications International Corporation, Torrance, California 

and 

Dennis P. Stocked and Sandra L. Olson* 

NASA Lewis Research Center, Cleveland, Ohio 



conduct in * 2 . 2 , NASA .ui te^^TtZTL^T* """^ ^ «»«<~ «« 
ignited in micro*™* environments Z, «, S T™ J • ^. d -"««*«c, of gM -j« dhTusto. ft*.,, 
g* in norm,! ^ lnd ^ J^T,. * SXti^JT* "" "^ "" 
different ignition methods result in different flame bS^Ij , *•■«»* *•• established. TV 

.djustment. and approach toward uSZ^J^^^lTT " reUti0 ° '° ""k^ *"—« 
studies, which werY in , «r JTn, ^ n^ed^d^T^' """ °< "* «-« »f the preriou, 
microgravity. ,„ .dditioa. I^Hb^JiZtS SfSll 2 r^^, *"' "*" ^ ta 
blue, whereas Mue was not discernible in the mTcro^JT * V"?**** (ntaed •» " *■—/■> were entirely 
B.«d on the* ^ „ ^^soT^^^^ <» *-«* 

Wue (and not visible on the high-speed Mm) fa. this ■» JTZ™ J!!T!L ^^^ "*■*• m, >' tav « »*■ 
observe of the flam, behavT^ dZj ^r.TJ.T.^tZ^ "" ,,,ed,0d, °' '" ,,d00 M- 
. ste.dy-st.te, parabolic mode. ha, shown sad*",™ .1—^^ STS^* P, * S " ttd - AM * C * ti °" rf 
Tor those flame, th.t reached . ne.~te.dy JFJ5%!ttZ££ " - *— "^ "*• 



Introduction 

T HE 'if-"? ' di, l usion flame " classifies those types of flames 

<£ rl*K '? thC fUd 3nd the ° X,dlZer are not pSni«d. S3 
he combustion process is governed by the diffusion of reac- 

tants toward each other. Fires are a complicated type of this 

class of flames Regardless of the nature of the fuel involved 

that' comrV?! 1 ' ° f l aSe0US) ' the fundam «tal mechanisms 
that control the combustion process are the same These 
include the coupled processes of mixing, chemical kinetics 
flame radiation, diffusion, inertia, soot Nation SdSJS 
non (in hydrocarbon flames), and, depending upon the 
Grashof or Froude number, buoyancy-induced convection 

nn Z' } i . T flameS are laminar or tu rbulent, depending 
on he relat,ve effects of inertia and viscous forces, which caJ 
best be desenbed in terms of the jet Reynolds number based 
km nL ^ Pr0pert K S ? d n0ZZ,e size F«nda™nul studies of 

o^urbu^Tff Sh ° U i d kad ,0 3n im P roVed »™leritanding 
of turbulent diffusion flames, which are of practical interest 
and are not fully understood. ' 

The behavior of most fires on Earth is dominated by 
buoyancy-mduced convection which masks the chemical- 



0645^2 a^aT™ a"* PreSe c" ted in P" 1 M ^ AIA A-88- 
U u .088 26,h Aerof P« &**» Meeting, Reno, NV. Jan 

a work of 8 ! h '! v "'° n ^« ved March 15, 1989. This paper is declared 
a work of the U.S. Government and is not subject to convrieht 
protection in the United States. copyngnt 

-Senior Scientist, Thermal Hydraulics Division. Member AIAA 
Author to whom correspondence should be sent. On-site at the NASA- 
Lewis Research Cemer. MS 500-217. Cleveland. OH 44135 

'Director of .he Combustion Science and Advanced Technolosv 
Department. Cha.sworth. CA. Currently at Rockwdl ImSSS 
Corpora,™. Rocketdyne Divis.on. Canog, Park. CA. MemS 

tNASA Project Scientist. 



ftSno f' t C,,0nS th3t are funda nicntal to the under- 
mem ^h K° mbUSt, ? n P hcnomena - '" low-gravity env.ron- 
ments the buoyant force is suppressed and the remaining 

cT*s and* klnet £' radiatiOD ' dlffuS,0n ' inert,a . *>" P'° 
^f« -^ i^ X ' ng b€COme res P° n «ble for the very different 
observed beha^or of laminar flames. As a result, ?s,de from 
he primary goal of understanding the behavior of fires m 
spacecraft environments, microgravity combustion studies fa- 
ahtate the study of the physico-chemical phenomena masked 
by buoyancy under normal-gravity conditions. In addScm 
numerical modeling of flames can benefit from the removal of 

XVomrwh^ ^" 1 ^ by f0CUS,n « ° n the re ™" "8 
phenomena, which are not separable and cannot be isolated 

Background 

Laminar diffusion names of hydrogen, methane, ethylene, 
and propylene have been studied'-' in the 2.2 s NASA Lewis 
Research Center Drop Tower. In these studies, the flames 
were ignited in normal gravity and allowed to reach steady 
state. The experiment package was then dropped, subjecting 
he flame to the microgravity condition. High-speed movies of 
the flames have shown that during a very short penod of 
adjustment from normal gravity to microgravity, a sudden 
decrease in the flame height (h) occurred within 0.05 s After 

V% Xe T> ° ^ flame height t0 a ™»i»"» wit" 
o.6< (WAl.X 1.0 (depending on the jet Reynolds num- 
ber), the name height increased with time. This resulted in 
either extinction with 1.0<M,.,)<1.3, approach toward 
an apparent steady state with 1.4<(A,_ f /*,.,)< 1.7, or con- 
tinual increase until the end of the test time Here, the sub- 
scripts mm. ex, »-g, and 1-j correspond, respectively, to 
minimum extinguished, microgravity. and normal-gravity 
flame heights. Compared to laminar names in normal gravity 
those observed in microgravity are larger, diffuse, and rather 
globular This is due to the absence of the buoyant convection 
leaving diffusion a much more important mechanism of 



FEBRUARY 1990 



LAMINAR GAS-JET DIFFUSION FLAMES IN M1CROGRAV1TY 



237 



2 9 



i 

i 
l 

% 



Extinction 



E 
u 



c 
— 

V 

E 
o 

"- 2.1 



1.7 



Fuel flow rite 

- 2.8 cc/s 

Nozzle radius 

- 0.186 ca 



1 

!> 

I Zero-f 



-0 4 



0.4 

Tim«, S 



8 



Fig. 1 Un«th«s«fiii»ctkMioftiiiielninkTor«i^ forMextta « u ^ 
methane-air diffusion flame. 1 



transport relative to the jet momentum. However, as will be 
seen later, jet momentum still affects the flame shape. 

The observed flame behavior in micrograwty, i.e., steady 
state, transient, or extinguished, has been attributed to the 
influence of the jet Reynolds number. Figures 1 and 2 show 
the behavior of two methane-air flames' 4 in micTogravity: an 
extinguished flame (Fig. 1) and a flame which apparenty 
reached steady state (Fig. 2) during the drop. The data 
presented in the previous works'" suggest that for a fixed tube 
size and variable fuel velocity, a certain range of the jet 
Reynolds number exists for which the 2.2 s available time is 
not sufficient to yield either flame extinction or steady state. 
However, it will be shown (see the Results and Discussion 
section) that a majority of these "transient" or even extin- 
guished" flames of the previous studies reached a near-steady 
state when the flame was ignited in microgravity. 

The reported'* microgravity flame behavior may best be 
described by the following mechanisms, based on the qualita- 
tive observations of the films. If the flame is ignited in normal 
gravity and then dropped, a rapid decrease in flame height 
occurs (see Figs. 1 and 2). This behavior is due to the sudden 
accumulation of the hot combustion products in the flame 
region, resulting from the loss of the buoyaacy-dnven connec- 
tive flow. The flame zone then becomes shielded from the 
oxygen, which promotes pyrolysis of the constantly flowing 
fresh fuel. A critical reduction in temperature occurs, caused 
by a reduced heat-release rate probably due to the change in 
the oxidation kinetics. This enhances soot formation and 
radiation cooling, resulting in the onset of a chemical-kinetic 
limitation on the neat release process. The combination of 
these effects and the jet momentum leads to either extinction 
or a slow readjustment to a steady-state flame (with convec- 
tive transport matching diffusive transport). 

The sensitivity of the extinguishment process to convective 
transport in microgravity has been demonstrated. Low 
forced-air velocities ( a 10 cm/s) in coaxial jets of methane-air 
diffusion flames were sufficient to sustain combustion in mi- 
crogravity, whereas similar microgravity flames in quiescent 
air reportedly extinguished. This indicates that relatively small 
forced-convection currents play a major role in combustion 
under microgravity conditions. 

The ignition approach previously used in the drop tower 
experiments exposes an established normal-gravity flame to * 
sudden change in the gravity level. It is not « :lear whether the 
reported microgravity flame behavior is due to this step 
change in gravity level or the microgravity environment lUelL 
In addition^ it is conceivable that buoyant forces could induce 



3.5 



E 

u 



£ 30 



ltote, — f • * • • 

zero-? • 

length 



E 
o 



2.5 



2.0 






.• • 



Fuel flo* rate 
■ 2.1 cc/s 

Nozzle radius 
- 0.051 ca 



-v: 

# Min. zero -9 lenflrn 



Ave. normol -g 
length 



Normal */' 



L_L 



Zero-? 

1 1 L 



• 04 00 0.4 08 1.2 1.6 2.0 

Time, s 



Fig 2 Lent* as a taction of time in microfrarity for a methane-air 
diffusion name that retched near-steady ittte. 4 




in. 1 Eneriaent enchase: a) combustion chamber, b) flow meter, c) 
^rt^Sfaa W* r««™ retulator, g) AD Une; h) coo*o 
bottle; k) to rent. 



r*- 



£ 




l^U 



Fta 4 Schematic dUcram of the flow system: a) combos** chamber 

hft h-Jr^-ate cTdumbtr bulkhead; d) l/tta. O.D. mbrng; .) 

] ^P^ ^^ Vernier .markinc* 1/4-- O.D. tubing; l) •> 

i^MI-d bcSe; m) hand «*«; .) q-ck dbcoooect; o) 



Tent*. 



238 



BAHADOR1, EDELMAN, STOCKER, AND OLSON 



AIAA JOURNAL 



Tible 1 Dili obtained from mkrognvity ind normal-gravity experiments' 



Flame 
number 



2 
3 
4 
5 
6 
7 
8 
9 
10 
II 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 



Nozzle 

radius, 

cm 



0.051 



0.0825 



0.051 



Fuel 



Methane 



0.051 Propane 

0.0825 



Methane 



0.0825 

0051 Propane 

0.0825 



Fuel volume 

flow rate, 

cm J /s 



10 

2.0 

3.0 

10 

2.0 

3.0 

0.5 

1.0 

1.5 

0.5 

1.0 

1.5 

10 

2.0 

3.0 

1.0 

20 

3.0 

0.5 

1.0 

1.5 

0.5 

1.0 

1.5 



Steady-state 

flame height, 

cm 



I.5 C 
3.5 
5.1 
2.4' 
3.7 C 
5.5 
4.0 
7.2 
111 
3.8 
7.3 
9.9 
09 
2.0 
4.0(7) 
0.9 
2.5 
3.6 
2.6 
5.6 
80 
3.0 
7.8(7) 
8.0 



S.S. maximum 

flame radius, 

cm 



067 

070 

0.75 

1.06 

1 05 

111 

090 

1.02 

111 

1.32 

153 

1.73 

0.35 

0.43 

0.36 

0.38 

0.60 

0.61 

0.45 

0.50 

0.50 

0.57 

0.50 

0.58 



'Question mirks indicate uncertainties in ihe measurements due to the dime dicker ~ ' 

Reynolds number is based on the nozzle radius. 
TTiese flames were d.nuse, faint, and resembling an undervenul.led (lame; it could not be concluded that they reached steady state. 



Approx. time 
to reach S.S., 



0.6 

0.8 

10 

1.0 

09 

16 

1.2 

0.8 

0.8 

1.2 

18 

1.3 

0.8 

13 

(7) 

0.6 

0.4 

0.3 

0.4 

0.4 

1.3 

0.4 

D 

0.4 



Reynolds 
number 6 



38 

76 

114 

24 

48 

72 

70 

140 

210 

44 

88 

132 

38 

76 

114 

24 

48 

72 

70 

140 

210 

44 

88 

132 



enough momentum during the pretest normal-gravity period 
to persist throughout the 2.2 s test time. In order to study the 
true characteristics of microgravity flames, it seems logical to 
simplify the initial conditions by igniting the flames in micro- 
gravity. The focus of this paper is to study the behavior of 
flames ignited in microgravity in order to achieve a more 
fundamental understanding of diffusion flames. 

Apparatus 

Laminar methane and propane diffusion flames in quiescent 
air were studied in the 2.2 s NASA-Lewis Research Center 
Drop Tower (facility described elsewhere 6 ). The experiment 
package and a schematic diagram of the flow system are 
shown in Figs. 3 and 4, respectively. The combustion chamber 
contains a burner, the lighting system, a viewing port for the 
camera, the spark ignitor, and ports for filling the chamber 
with oxidizer and for venting the burned gases. The other 
components of the system are shown in Fig. 4. The volume of 
the combustion chamber is approximately 0.04 m J . Two fuel 
nozzles with inside radii of 0.051 cm and 0.0825 cm were used 
in the studies. The fuel system was calibrated, based on the 
standard conditions of 0°C and 1 atm. The fuel flow rates 
were 10, 2.0, and 3.0cm 3 /s for methane, and 0.5, 1.0, and 
1.5 cm /s for propane. The 16 mm movie camera was origi- 
nally set for a filming rate of 30 frames/s and the film (Tung- 
sten Eastman Ektachrome Video News) was force-processed 2 
f-stops. At this filming rate, the dim blue portion of the 
microgravity flames was not clearly visible in the processed 
films. Visibility was subsequently improved by reducing the 
framing rate to 15 frames/s. 

A spark electrode was located 4 mm above the nozzle tip 
and 4 mm away from the nozzle centerline. The electrode was 
perpendicular to the centerline of the camera. The nozzle, 
electrically grounded, acted as the second electrode. The spark 
was pulsed at 4.8 Hz, with approximately 18 mJ/pulse. The 
flames were ignited during the first 0.2-0.4 s in microgravity. 
For all of the tests, the spark-ignition system was active for 
the entire 2.2 s test time and about 0.8 s after impact. The 
ignitor occasionally did not arc to the burner tip but shorted 



out elsewhere, however, this did not affect the flame behavior 
A frame-by-frame study of the movies showed no disturbance 
of the flames by the spark during the 3.0 s of burning. 

The 500-cm 3 fuel bottle contained either methane or 
propane filled to 2 atm. The dry air in the combustion cham- 
ber was composed of 21% oxygen and 79% nitrogen, and had 
a pressure of 1.0 atm in the chamber. Due to the short 
duration of each test, the amount of oxygen consumed was 
negligible, always less than 0.3% per flame. The air in the 
experiment chamber was changed after a number of tests were 
conducted, when < 5% of oxygen was depleted. Details of 
the experimental procedure can be found elsewhere. 10 

Results and Discussion 

Table 1 shows the experimental conditions and the average 
steady-state flame lengths, maximum flame radii, and approx- 
imate time to reach steady state for both normal-gravity and 
microgravity flames of methane and propane. The normal- 
gravity flames flickered (except at low-flow rates), and the 
steady-state flame heights and maximum radii are averages. 
Figures 5 and 6 show the flame lengths as a function of time 
for methane and propane flames in microgravity, respectively. 
Unlike flames in normal gravity, those under the microgravity 
conditions are fairly steady and do not flicker. The methane 
flames at low volume-flow rates were very faint and appeared 
to have an open tip. They were sufficiently visible to identify 
the maximum radius, but beyond this height, visible radiation 
could not be observed and the flame appeared to be underven- 
tilated. This is shown by the truncated data given in Fig. 5. 
The higher flow-rate methane flames reached near-steady state 
with some fluctuations at the later times. Previous studies of 
these methane flames with the 0.0825 cm nozzle radius have 
shown a transient state at the completion of the drop. 4 In 
general. Table 1 shows the following relationships for flame 
heights (A) and maximum flame radii (r„): 1.2 <(/»_/ 
h t _,) < 1.8 for both nozzle sizes; 1.6 < {T m ^_Jr m , ) < 2*2 
for the smaller nozzle; 2.3 <(r mjt .Jr mA _,) < 3.1 for the 
larger nozzle. Here, the subscripts n - g and 1 - g correspond 
to the microgravity and normal-gravity flame dimensions, 



FEBRUARY 1990 



LAMINAR GAS-JET DIFFUSION FLAMES IN M1CROGRAVITY 



239 



respectively An overshoot in the methane flame heights can 
be seen in Fig 5, whereas the propane flames in Fig. 6 do not 
show this behavior. This observation is currently under inves- 

Uga Figure 7 shows a comparison between the two methods of 
ignition for a methane flame, with a fuel flow rate of 3.0 cm /s 
(2 9cm 3 /s for the normal-gravity ignition case ) and a nozzle 
radius of 0.051 cm. The trends in flame development are 
similar, and the steady-state flame heights are the same. 
Typical normal-gravity flame flicker is also shown in Fig. 7. 
The normal-gravity ignition method suggests that the micro- 
gravity flame behavior may be affected by the flickering 
normal-gravity flame height at the instant the package is 
subjected to the microgravity condition. 



The flame heights observed in the previous and present 
studies are compared in Fig. 8. The jet Reynolds number is 
selected as a common parameter because it represents the 
effects of nozzle size, fuel type, and jet momentum, which are 
shown to control the behav.or of the flame as far as extinction 
and approach toward steady state are concerned. In Fig. 8, 
the mangles represent those flames which have reached a 
(near-) steady state. 4 The flames which are reported extin- 
guished are also shown (circles); the heights correspond to 
fhose at extinction during the drop. To distinguish between 
the different regimes, lines are drawn to connect the different 
available data points. Thus, the solid lines cover both the 
steady-state and extinguished regimes for which reported data 
is available. The broken line for the larger nozzle shows the 




1 ' * 4 ' Jl ' & ' *'« ' ^ * U ' ig ' h ' U 

Tlm« (««c) — — 

Fig. 5 Mkrogravity methane Ban* heights. 



12r> 
11 

10 

t • 

8 

I* 

u 

h 



* 3 
ik 

2 
1 



o o 



o o 



o o O o o 



A A 



A A 



• o 



A A 



* A 



A A 

A 
A 



*a . . ■ S 5 B 3 B B B 

■ ! o ° u 



■ o 
■ a 
a 



i ■ ■ * 



iy*\ 



End of 2.2 
Second Drvo 



■ 
O 
A 
a 



Mill* 

Wlni(ci) 



0.0S1 



0.0SM 



Fuel Mm 
UU(c»V») 



l.i 

1.0 
0.J 

l.» 

1.0 

o.s 



0.4 



' ■ - ' - ' - ■ a U ' 2.8 ' 3.2 ' 3.6 



08^ 1.2 1.6 % 2.0 

Tlm« (mc)— ' 



Fig. 6 Mkrofra»hy propane torn* height*. 



BAttADOKl, hUhLMAN, STOCKER, AND OLSON 



6.0 



3 
u 



4.0 - 



• • 



to<*P c *»oaiocyiooA D %3t3^ » 



j; 



2.0 



a* B 



□ 

a 
o • 



Zero - 7 



a NORMAL-GRAVITY ICNITION 
• LOW-GRAVITY IGNITION 



J_ 



J. 



-.4 0.0 0.4 0.8 1.2 

TIME. SECONOS 



1.6 



2.0 



2.* 



Fig. 7 Comparison between the normal-gravity 4 and uicnwravifv 
igiuaon of dame No. 6 of Table 1. 



AIAA JOURNAL 



range of Reynolds numbers for which the flames required 
longer than 2.2 s to establish their final state. This regime 
comprises the -'transient" flames of the previous work 4 for 
which no heights are available. In addition the dotted' lines 
correspond to those regions for which no flames were studied 
Therefore the dotted and broken lines are not meant to 
indicate that flame heights can be deduced at intermediate 
Reynolds numbers from Fig. 8. Rather, they g,ve the ranges 
ot Reynolds numbers which cover this class of flames The 
squares shown in Fig. 8 are the observed (near-) steady-state 
name heights of the present study, obtained from Fig 5 
Although the truncated data in Fig. 5 for the lower flow rates 
may give the impression that the flames extinguished these 
flames indeed existed throughout the drop. As mentioned 
before they were very faint and appeared to have open tips 
with indeterminate length, resembling typical underventilated 
name behavior. 

«? Ca \ b f« SCCn '" Fig ' 8 ' that the P« v '0"sly reported 
extinguished flames (circles) may not extinguish if the flame is 
ignited in microgravity. However, low-flow rate methane 



20 
1* 

8 " 






S • 



Tr«uu„t flaws (no (Might avail able) 

.... Ho data available ' 
0«u avallablt 

A Steady-state f lanes 

• Extinguish** flaws 



• O Present study 
(seeFle.j) 



9..- 




Nozzle Radius. 



* ~-T~ l I — I — I.I I I ■ ■ ■ I- 



» «0 M 80 100 120 140 ' 



Reynolds Nuaber » 

Fig. I Microgravity metbane-flame heights as a function of jet Reynolds 



160 180 200 



number (see the text for details). 



12 



10 



Exp. 



Theory 



Fuel 

CH4 

CH4 



r 
(«) 



0.051 
0.0825 




_L 



30 W 90 120 ISO 180 210 240 

Reynolds Number 
Fig. 9 Predicted and measured microgravity methane-flame heights vs Jet Reynolds number. 



FEBRUARY 1990 



LAMINAR GAS-JET DIFFUSION FLAMES IN MICROGRAVITY 



241 




Fig. 10 Predicted »nd 



mewured micro«r.vit, prop.n-n.me hei.Ms r* Jet Reynolds number. 



flames in the present study (filmed at 15 frames/*) ^were 
entirely blue, whereas blue was not discernible in the micro 
SS flames of the previous works (filmed at 400 frames/s). 
K on these findings, it is concluded that the previously 
reported "extinguished" flames may have been blue (and not 
visible on the high-speed films). ,,„■„„,,. 

A previously developed computational model is used to 
predict steady-state flame heights for comparison with the 
experimental data. The parabolic model for studying laminar 
gasket diffusion flames under arbitrary gravitational accelera- 
tions consists of the conservation equations for elements, 
S momentum, and energy. It includes the effects of iner- 
tia viscosity, diffusion, and chemical reactions. The chemistry 
assumed is "that of shifting ^^^^^^Zt 
multicomponent diffusion are considered The radiation 
model is a thin-gas approximauon using carbon dicjideand 
water vapor. The differencing scheme is an explicit finite 
difference technique. The results have shown a very good 
agreement with the earlier experimental data for both normal- 
Sand microgravity condition,.*' However, it was 
Sown that accurate flame shape predictions require an accu- 
rate prediction of the relative rates of molecular transport of 
Snes momentum, and energy. This steady-state, parabolic 
model has been applied to the "«^V *?« **£ [ 
and 6 The results are shown in Figs. 9 and 10, where a 
iact™ trend in the predicted behavior of the Aam« is 
demonstrated. However, the propane names are Jg£d 
longer and the methane flames shorter than the predicted 
Ses One possible explanation is the absence of soot and 
fhTSsoaated^radiation in the model i»^ "Jjjjg 
concentrations of soot observed in most of the mcrogravity 
flame7 n addition, the absence of elliptic effects . (««l diSu- 
SoMnd £ relatively simple predictions for the molecular 
ransport rates (i.e., species, momentum, and energy) may 
contSeTo the differences between the predicted and mea- 

sured flame heights. ...^, , h - rtret i- 

Figure 11 shows a comparison between different theoreti 
cal and experimenul results. The data points correspond to 
Se steady-state flame heights of the present study and previ- 
ous wort ; « The predictions are those presented in F.gsJ and 
?f - WO ';., .. !L r«.,lM of another theoretical modeling 




Nil 


0«u; 
tafl 4 
UX S 


Oiu; 

Prttwit 

Study 


IWtMM 


A 


▲ 


(M>rif*« 


o 




frsoy' •"• 


a 




rnftf 




♦ 



*id1«1oiit (««f. II) 

"~ \ »r«dlctlon> (pf«n«t Jtudrl 



00 200 300 

Reynolds number 



ioo 



«f fad* reproduced." w«h modincrioo. ud .ddWo-. 



BAHADORI, EDELMAN, STOCKER, AND OLSON 



AIAA JOURNAL 







effort. The two numerical models show a good agreement 
with each other and with the data. The propylene-flame 
heights of the previous studies 4 do not agree with the predic- 
tions. This is surprising, since propylene and propane are 
very close in chemical structure and satisfactory agreement is 
obtained for methane, ethylene, and propane. 

In gas-jet diffusion flames, the heat-release mechanism is 
not uniform throughout the flowfield. Pyrolysis, producing 
soot and partially oxidized species, dominates in the near-jet 



region, while the continued heating and increased residence 
time result in the tendency to burn off these species down- 
stream in the vicinity of the flame tip. However, pyrolysis and 
soot formation reduce the temperature level in the flame and, 
with the aid of the increased residence time, sooting is en- 
hanced. This, in turn, results in enhanced radiation, more 
cooling, and delayed soot burn-off downstream of the flame. 
This effect is much more pronounced in microgravity flames 
than those in normal gravity, due to the lack of buoyancy. As 



FEBRUARY 1990 



LAMINAR GAS-JET DIFFUSION FLAMES IN MICROGRAVITY 



243 




Fig. 12 Continued. 



a result, thcrmophoretic effects may become an important 
factor in the regions of large temperature gradients with high 
soot concentrations. The observed open tip of the propane 
flames and quenching at the tip give the appearance of an 
underventilated flame (see Fig. 12). 

In Fig. 12, a frame-by-frame analysis of flame No. 11 of 
Table 1 is presented. This propane flame reaches steady state 
and remains fairly uniform throughout the drop. The colors 
indicate the presence of particular species in hydrocarbon 
flames, e.g., blue-violet region (outer region at the base 
showing the spectrum of burning CO) and bluish-green region 
(showing C 2 and CH emissions, where C 2 bands appear green 
shaded toward violet and CH bands appear violet-blue). A 
highly luminous zone (bright yellow) indicates the thermal 
emission of burning carbon particles. This luminous zone 
becomes yellow, then orange, and then red and dark red 
toward the boundary of the visible region, as the temperature 
of the unbumed soot decreases. The reddish-orange color of 
the flame is due to the soot at cooler temperatures than would 
be expected for typical burnoff conditions. 

The color of the radiating soot, in conjunction with Wien's 
law, can be used as an indicator of the approximate local 
flame temperature. The cloud of burning soot particles is 
yellow, i.e., ~0.58-/*m wavelength (1370 K < T < 1670 K, 
where 1670 K is the temperature of a white hot body). As the 
temperature decreases, the soot particles become orange red 
(~1370K ~0.6/im), then cherry red (~ 1 170 K., 0.6- 
0.65 /im), and finally dull red ( -970 K, 0.65-0.70 /im). These 
effects are shown clearly in the results of Fig. 12, which 
indicate cooling by pyrolysis and enhanced radiation due to 
increased soot formation. 

Frame number 35 of Fig. 12 shows the flame during the 
rapid deceleration as the experiment package is brought to 
rest at the end of the drop. The high level of induced 
convection causes blowoff within <0.l s After the impact, 
reignition can be seen due to the spark pulses, resulting in 
normal-gravity behavior, i.e., flame flicker. Examination of 
high-speed movies for the deceleration period reveals the 
behavior of the flame in the transient state under rapid 
changes in gravity level, providing byproduct data for further 
analysis. A study of laminar diffusion flames under elevated 
gravitational conditions is reported elsewhere. 12 

Conclusions 

Ignition of laminar methane and propane diffusion flames 
in microgravity environments has been demonstrated. The 
new ignition method combined with the refined photographic 
technique has shown that flames previously thought to have 
been extinguished were still burning during the entire 2.2 s of 
microgravity. 

The low-flow rate methane flames are blue and very dim, 
prompting further studies to determine the extinction limits 
for laminar diffusion flames in microgravity environments. 

Many flames (especially those of propane) have an open tip 
resembling an underventilated behavior. This is attributed to 
radiation, soot formation, and possibly thermophoretic 
effects, which appear to become more important in flames 
under microgravity conditions. However, a thorough analysis 
must be made to determine the relative importance of each. 

The predicted flame heights, using a steady-state parabolic 
model, are in good agreement with the measured steady-state 
flame heights. 

Longer test times are planned through the use of the 5.18 s 
Zero-Gravity Facility and the KC-135 aircraft (free-floating, 
10-15 s). Flame visualization, and radiation, temperature, 
and concentration measurements at these longer test times are 
going to provide a more complete quantitative description of 
these flames. 

Acknowledgments 

This work is supported by NASA Lewis Research Center 
under Contract NAS3-22822. The authors would like to 



244 



BAHADORI, EDELMAN, STOCKER, AND OLSON 



AIAA JOURNAL 



express their appreciation to Professor Irvin Glassman for a 
discussion on the interpretation of the colors observed in 
microgravity flames. They also wish to thank Mr. Matthew 
Hart for conducting the design-feasibility drop tests, and Mr. 
Thomas Morrissey for conducting parts of the numerical 
calculations. 



References 

'Cochran, T H. and Masica, W J , "Effects of Gravity on Laminar 
Gas Jet Diffusion Flames," NASA TN D-5872, 1970. 

: Cochran, T. H. and Masica. W. J., "An Investigation of Gravity 
Effects on Laminar Gas-Jet Diffusion Flames," Thirteenth Symposium 
(International) on Combustion, The Combustion Inst., Pittsburgh PA 
1971, pp. 821-829. 

'Haggard, J. B. Jr. and Cochran, T. H„ "Stable Hydrocarbon 
Diffusion Flames in a Weightless Environment," Combustion Science 
and Technology, Vol. 5, Aug. 1972, pp. 291-298. 

'Cochran, T. H., "Experimental Investigation of Laminar Gas Jet 
Diffusion Flames in Zero Gravity," NASA TN D-6523, 1972. 

'Haggard, J. B. Jr. and Cochran. T. H., "Hydrogen and Hydro- 
carbon Diffusion Flames in a Weightless Environment," NASA TN 
D-7165, 1973. 

'Edetman, R. B., Fortune, O. F., Weilerstein, G., Cochran, T. H.. 



and Haggard. J. B Jr. "An Analvtical and Experimental Invesiigation 
of Gravity Effects Upon Laminar Gas Jet Diffusion Flames." Four- 
teenth Symposium (International) on Combustion, The Combustion 
Inst , Pittsburgh, PA. 1973, pp. 399-412. 

'Edelman, R. B, Fortune, O.. and Weilerstein, G . "Analytical 
Study of Gravity Effects on Laminar Diffusion Flames " NASA 
CR-I20921, 1973. 

'Haggard, J. B Jr , "Forced and Natural Convection in Laminar- 
Jet Diffusion Flames," NASA TP 1841, 1981. 

'Edelman, R. B. and Bahadon. M Y , "Effects of Buoyancy on 
Gas-Jet Diffusion Flames: Experiment and Theory," Acta Astronau- 
nca. Vol. 13, No 11/12, 1986, pp 681-688. 

l0 Edelman, R. B . Bahadon. M. Y„ Olson, S L., and Stocker, D. 
P.. "Laminar Diffusion Flames Under Micro-Gravity Conditions " 
Paper AIAA 88-0645, AIAA 26th Aerospace Sciences Meetine 
Reno, NV, Jan. 1988. 

"Klajn, M. and Oppenheim, A. K , "Influence of Exothermicity on 
the Shape of a Diffusion Flame," Nineteenth Symposium (Interna- 
tional) on Combustion, The Combustion Inst., Pittsburgh PA 1982 
pp. 223-235. 

l2 Altenkirch. R. A., Eichhorn, R., Hsu, N. N., Brancic, A. B., and 
Cevallos, N. E., "Characteristics of Laminar Gas Jet Diffusion 
Flames Under the Influence of Elevated Gravity," Sixteenth Sympo- 
sium (International) on Combustion, The Combustion Inst Pitts- 
burgh, PA, 1976, pp. 1165-1174. 



Section 1Q 



"AN ANALYTICAL SOLUTION FOR TRANSIENT, CYLINDRICALLY SYMMETRIC 
LAMINAR DIFFUSION FLAMES IN THE ABSENCE OF BUOYANCY" 



M. Y. Bahadori 



Paper presented at the Central States Meeting of The Combustion Institute, 

Cincinnati, Ohio, May 1990 



An Analytical Solution For Transient, Cy lindrically Symmetric 
Laminar Diffusion Flames in the Absence o f Buoyancy* 

M. Yousef Bahadori 

Science Applications International Corporation 

Torrance, California 

Abstract - Analytical solutions are obtained, to the first approximation, for 
the transient shape and height of overventilated, cylindrically symmetric, 
laminar diffusion flames in the absence of gravity. Equal and constant flow 
speeds, transport parameters, and physical properties are assumed for the fuel 
and oxidizer streams. Axial and radial diffusion are accounted in the 
formulation. The solution yields the development of the flame boundary from 
ignition toward steady state, and reduces smoothly to the classical solution 
of laminar diffusion flames as-, independently, axial diffusion becomes 
negligible, and steady state is approached. Comparisons with the results of 
microgravity experiments show a satisfactory agreement for the increase in 
flame height with time. The solution is useful in estimating the time 
required for the development of the flame toward steady state following 
ignition in microgravity environments. 

1. Introduction 

This analysis is based on, and is an extension of, studies on heat flow 
in composite cylinders [1], and on defining the boundaries of cylindrically 
symmetric laminar diffusion flames with equal flow velocities and diffusion 
coefficients in the absence of buoyancy [2]. Recently, analytical solutions 
have been developed for flame shapes with arbitrary values of diffusion 
coefficients and inlet velocities of the fuel and oxidizer [3]. Also, the 
Burke-Schumann model [2] has been extended for two competing laminar diffusion 
flames with cylindrical symmetry [4], and two and three flames with 
rectangular symmetry [5] . 

Efforts to study laminar diffusion flames under near-zero-gravity 
conditions have so far been limited to 2.2 seconds of microgravity in the 
NASA-Lewis Drop Tower. In these experiments, a self-contained gaseous 
diffusion-flame apparatus was allowed to fall inside a falling drag shield. 
The method provided time-resolved data on the development of flame shape for 
the near-zero-gravity condition achieved during the drop. Various hydrocarbon 
diffusion flames burning in quiescent air have been studied [6-13], and some 
of these works have been reviewed in a recent article [14]. Currently, 
preliminary studies of microgravity laminar diffusion flames with secondary 
air flows are underway [15], with the intent of comparing the results with 
Burke-Schumann type models [2,3]. 

♦Presented at the Central States Section Meeting (Spring, 1990) of The 
Combustion Institute. Cincinnati, Ohio. May 20-22, 1990. This work is 
supported by NASA Lewis Research Center under Contract NAS3-22822. 

f Senior Scientist; on-site at NASA Lewis Research Center. MS 500-217, 
Cleveland, Ohio 44135. 



2. Formulation 

In this section, a problem is formulated for overventilated, laminar, 
transient, cylindrically symmetric diffusion flames to obtain, to the first 
approximation, a closed-form mathematical representation for the development 
of the flame boundary with time following ignition in a zero-gravity 
environment. The assumptions used in the formulation are as follows: (a) 
constant pressure, (b) unity Lewis number, (c) equal and constant flow 
velocities, transport parameters, and physical properties in the fuel and 
oxidizer streams, (d) flame-sheet approximation, and (e) no buoyancy effects 
Axial diffusion is included in the formulation. 

For the cylindrically symmetric flow of gases, we consider the model of 
Fig. 1, where an inert gas flows in the inner of two concentric cylinders for 
times t < 0. The oxidizer flows in the outer cylindrical shell at all times. 
The flow of the inert gas is replaced by a fuel flow of the same velocity at 
t = 0, i.e., the interface between the fuel and inert gas arrives at the 
burner port at t = 0, and ignition occurs at this instant of time. 

The species-conservation equations for the fuel and oxidizer above the 
burner port at times t>0 reduce to a single differential equation 

ft^fz'^TTr'Tl'—z) *r J<r<d72. « > 0. t > 0, 

or 3z 

once we impose the conditions u F = u = u = const, and D F = D = D 
in addition to introducing a single variable 



(i: 



const 



1Y F for 4 r < r f , z > 0, t > 0, 
-vY Q for r f < r < d'/2, z > 0, t > 



(2) 



Here, Y is the mass fraction, u is the velocity, D is the diffusion coeffi- 
cient, t is the time, z is the axial coordinate, r is the radial coordinate, 
r f (z,t) is the flame radius, v iz the stoichiometric ratio of gram.. o£ fi el 
consumed per gram J oxidizer at the flame front, and subscripts F and 
denote the fuel and oxidizer, respectively. 



ylnert- 
/ oxidizer 
boundary 



d'/2 

Inert-fuel 
boundary 



// 
i < 



Flame 



r f U,t) 



n 



rr 



uz 



• / boundary 
at tine 

t 2> t l 



-Fl ame 
boundary 
at t1«e 
tj> 



d'/2 



Fig. 1. 



(a) t<0 



(b) t>0 



Schematic diagram of concentric cylinder* with Inert I entering at 
z-0 for 0<r<d/2 during the tlaei t<0 [see (a) J. which is rep'eced 
by the flow of fuel P at t£0 [see (b)]. The oxidizer Is entering 
at z-0 for d/2<r<d'/2 at all tlaes [(a) and (b)]. 



In view of Eq. (2), the initial conditions (denoted by subscript o) are 

10 for < r < d/2, z > 0, t = , 
- vY n « for d/2 < r < d'/2, z > 0, t « 0, 
Y_ for u < d/2, z ■ 0, t > 0, 



IY- for u < d/2, z 
-vY Q for d/2 < r < d'/2, 



Y(r,0,t) = (4) 

Y Q for d/2 < r < d'/2, z - 0, t > 0. 

The symmetry and boundary conditions are, respectively, 

3Y(0,z,t)/3r - 0, (5) 

3Y(d72,z,t)/3r - 0. (6) 

The flame-sheet approximation defines the location of the flame boundary, viz. 

Y(r f ,z,t) - 0. (7) 

We now derive an expression for the flame shape in the form f(rf.z.t) = 0. 
3. Method of Solution 

Using separation of variables, the general solution becomes 



Y(r.x.t) - t t E jk VY 5 ex P[' D(X J- ^ H ] "'[W 1 ->/l + ^ k D 2 /u 2 )z] . (8) 

where the Ej k are constants, J is the Bessel function of the first kind of 
order zero, and Xj and /?j k are eigenvalues. In deriving Eq. (8), the 
symmetry condition has been used to show that Bessel functions of the second 
kind of order zero can not exist. In addition, the boundary condition yields 

J 1 (X j d'/2) - . (9) 

which is the equation determining the eigenvalues Xj . Here, J 1 is the Bessel 
function of the first kind of order one. 

Using the initial conditions, the coefficients Ej k and eigenvalues /?j k 
can be evaluated. Substitution of Eqs . (3) and (4) in Eq. (8) yields 

«— — ■ i f for 0<r< d/2. z > 0. t - 0. 

E Jo (X j r)E Ejk e X p[^(l- > /r776^W)zJ.| _^ ^ d/2<r<d>/2> 2>Q> t . , dO) 

r , , , , i f *F o for « r < d/2. z • 0. t > 0. 

£ J n (X ir ) £ E, k exp|-D(x2 . 2 W . • (11) 

j-1 ° J k-1 Jk L V J jk/ J I -vY n „ for d/2 < r < d'/2. I - 0. t > 0. 

We hold z constant and vary r over the range of the problem in Eq. (10). The 
inner sum is therefore a constant depending on j, and denoted by Gj(z). Hence, 

» [ for < r < d/2, z > 0, t « 0, 

£ GAz) Jn^r) - | /«,» 

j-1 J J I -vY Qo for d/2 < r < d'/2. z > 0, t - 0. u "' 

Multiplying both sides of Eq. (12) by rJ (Xjr)dr, integrating over the range 



0^r<d'/2. and using the orthogonality properties of Bessel functions, it cap 
be shown that for Xj*0(i.e.,j>i), 

2vY Q (d/2) J.(A.d/2) 

G J (Z) = 2 2 ' for J ->1 > d3) 

J X j (d'/2) 2 jJ(X j d'/2) 

Eq. (9) has been used in the derivation of Eq. (13). The approach for G,(z) 
needs special treatment involving expansions of Bessel functions and taking 
the limits as Xi—0, which ultimately gives 

G l (z > " -^O.oD - (d/d') 2 ] . for j - 1. (14) 

Hence, substitution of Eqs. (13) and (14) in Eq. (10) yields 

£ E lk ex P[lD I 1 -V 1 + 40 lk° 2 / u2 ) z ] ■ -vr 0i0 I 1 - « d / d ') 2 ] • f °r J-l. (15) 

We now hold t constant and vary r over the range of the problem in Eq. (ll). 
Again, the inner sum is a constant depending on j. Following the same 
approach in deriving Eqs. (15) and (16), it can be shown that 

00 

£ E lk exp (D0 2 k t) -(Y F , o +vY 0>o )(d/d-) 2 -vY 0>o . for J - 1. (i 7) 

"-» Jk L J Jk J A j( d72) 2 J 2 (X jd 72) * ^ j * U <"> 

The flame location (where Y - 0) is obtained from Eq. (8), viz. 

t J (Xjr f ) exp (-DA 2 t) t E jk exp (OB^t) exp[^(l -fiT7ffi?)z] 

- -£ E i k ex > <<<> «p[fe(» -V^IA 2 )«] • (19) 

Equations (15)-(18) give the coefficients Ej k and eigenvalues /3 Jk (j£l). 

So far, we have not deviated from, or made approximations in the rigorous 
approach for obtaining the solution. However, Eqs. (15)-(18) show that for 
each Xj, there may exist an infinite number of coefficients E Jk and eigen- 
values /? jk . This makes the task of analytically examining the solution 
impossible. We therefore make the following assumption to obtain a first- 



order, approximate solution : At any fixed z and t. there exist only one E jp 
and one 0* y for each X j . The assumption may be justified due to the fact 
that the classical solution [2], using only the first nonzero eigenvalue, has 



given satisfactory results for normal -gravity flame heights [16], especially 
in the light of the restrictive approximations made on the physical grounds to 
formulate the original problem. With this assumption, Eq. (19) reduces to 



j? 2 E J VW exp [- 0U J - 8 j 2)t ] M "[lD (» ->/l*40jV/u 2 )z] 

A -Ej exp(O0*t) txp[Jjjj(l -^1 ♦ 48 Z D 2 /u 2 )z] , 



(20; 



and Eqs. (15)-(18) yield 
El exp[^-(l ->/l + 40 2 dV)z]* -vY 0fO [l - (d/d 1 ) 2 ] , for J-l. 



2vY n (d/2) J,(X d/2) 

2ofi , — -i— J . for J > 1. 



E, exp [— (l -Jl ♦ 4bV/u 2 )z ■ — t~T. T ' 

J H L2D v ' J ' J X.(d'/2) Z jJ(Xjd72) 

El exp(Deft) - (* F>0 *vY 0f0 )(d/d') 2 -vY 0> 

r r n,>2 « 2 ul • 2 ( Y F.O >VY 0,o) (d/2)J l ( V /2) 

E, exp -DU( - 6i)t ■ ! 5—57 " • 

J L j J -I Xj(d72) Z JJ(X d'/2) 

Proper manipulation of Eqs. (21)-(24) for the four unknowns, and substitution 
in Eq. (20) yields, finally, 



for J-l, 



for J > 1. 



(21) 

(22) 
(23) 
(24) 



E 2 ( Y F Q * vY n J (d/2) J l (X i d/2) 
— i-E«2 Mil LJ j ( X r ) exp 

j. 2 Xj(d72) 2 J§(Xjd72) ° J f 



u(t + z/u) . 



- [( Y F.o + vY 0,o) (d/d '» 2 * vY 0.o] "P^TBt 




40t 



PA 2 t ♦ T» (1 ♦ Y F>o /vY 0iO ) 
u 2 (t ♦ z/u) 2 



■I 



r Y. ft (d/d') 2 -i 

L vY ,o[l - (d/d') 2 ]. 



u 2 (t ♦ z/u) 2 



(25) 



which is the equation for flame shape. The flame height can be obtaind by 
setting r f » in Eq. (25). 

In the limit of t— <a , Eq. (25) yields the solution to the steady-state 
problem with axial diffusion [3], i.e., 

£? Xj(d72) 2 JJtXjd'/Z) 0jf L 2D \ ^ u Z / J 



[(v.o + vr 0iO )(d/d') 2 -vY 0fO ] 



(26) 



In the limit of (2DXj/u) 2 «l (i.e., negligible axial diffusion), Eq. (26) 
reduces to the classical solution of Burke and Schumann [2,3], as follows: 

>2, 



^ 2 (y p +vY 0o )(d/2) J.(X f d/2) / »ji\ 

V* L_L2 2i2i- LJ J X,r f exp ( - —J— 1 

4j Xj(d72) Z jg(Xjd72) ° Jf V " ' 

•-[( Y F.o + ^0.o)^ d '^- vY 0.o] 



(27] 



It should be noted that Eq. (25) upon using 



Y^ . (d/d') 2 



4Dt 1 



and 



I vY.Jl.(d/d') 2 jJ 



4Dt[pX 2 t ♦ In (l ♦ Y F>o /vY 0tO )] 

u 2 (t ♦ z/u) 2 
yields the transient solution in the absence of axial diffusion, viz. 



O.o 

u 2 (t ♦ z/u) z 



«1 . 



E 

i'2 



2 ( Y F,o+ vY 0,o) «'V Jl^j^) 



Ajtd'/Z) 2 J§(Xjd72) 



W 



f ) exp |- 



[oxjt + In (1 ♦ Y F>o /vY 0>o )] 



-[( Y F.o + ^ ,o)( d / d ') 2 -^O.o] exp 



In 



u (t + z/u) 
»F.o ( d / d '> 2 



vY, 



0,o 



1 - (d/d')2 



U (t + 2/u) 



(28) 



This expression can also be obtained by formulating the problem posed in Sec. 
2 without axial diffusion, and following the same method of solution outlined 
in the present section. 

4. Results and Discussion 

Equation (25) was solved numerically using polynomial expressions for 
Bessel functions. Up to 2000 -eigenvalues were calculated in the program using 
Eq. (9). A large number of eigenvalues was needed (especially at earlier 
times) due to the slow decay of the exponential terms. The flame shape was 
salved for selected times. This is similar to solving the classical solution 
for the flame radial location at any fixed height. 

Figure 2 shows the comparisons for two microgravity flames: an extin- 
guished flame [6], and one which reached a near-steady state [7] in 2.2 seconds 
of microgravity. In these experiments, the flames have been ignited in normal 
gravity, and after attaining steady state, were subjected to a microgravity 
condition. As a result, the flame height drops to a minimum upon release of 
the package due to the consequent entrapment of the combustion products caused 
by the removal of buoyancy, and then starts to develop due to the continuous 
flow of fuel (see Fig. 2). The flame of Fig. 2(a) shows extinction, which the 
present model is not capable of predicting; however, recent studies [11] have 
shown that by Igniting the flame In microgravity and improving the photographic 
technique, flames which were previously reported as extinguished [7] were 
probably burning during the entire 2.2 seconds. The intent In this paper is 
to investigate the behavior of the analytical solution, and make preliminary 
comparisons with some experimental results on flame development in general. 
Work is underway to apply the model to more recent data, e.g., those of Fig. 3 
which show flame development following ignition in microgravity. It is 
anticipated that similar results will be obtained, especially since 
comparisons between normal -gravity and microgravity ignition (see Fig. 4) have 
shown that the trends in flame development are similar, and the flames reach 
the same steady-state height. Note that although the experimental results of 
Fig. 2 correspond to a quiescent oxidizing environment, the model which is 
formulated for co-flowing configurations can predict the flame-height develop- 
ment fairly accurately through selection of an appropriate diffusion coeffi- 
cient, especially in the light of the restrictive assumptions made to formulate 
the problem (Sec. 2). Since the diffusion coefficient is the sole controlling 
parameter in the Burke-Schumann type solutions, Fig. 2 also shows the 
parametric effect of diffusion coefficient on flame-height behavior, confirming 
that the larger the diffusion coefficient, the shorter the flame height. 

Figure 5 shows the predicted development of shape with time for a 
methane-air flame. Note that the predicted maximum-flame radii are smaller 
than the experimental results (not presented here) due to the presence of air 




• o.a aft\ 



4.0 - 



3.0 



,. „,,. ,-, Ml Zm 



2 

i 



2.0 




1.5 3.0 



0.S 



CiptHatnt 
Trtfultnt «xf«l 
Theoretical S.S. 



h„ ■ 4.18 oi 

™ 3 — McV 



3.35 



«j • 0.OS1 ca 
g • 2»1.2 ca/i 
J I I 



1.3 



l.S 

Tlw, < 



1.0 



Fig. 2. Comparisons between the analytical and experimental variations of 

flame height with time. The experimental results show flame extinc- 
tion at approximately 0.6 sec [Fig. 1(a)], and reaching an apparent 
steady state in 0.4 sec [Fig. 1(b)]. The binary diffusion coeffi- 
cient of methane in nitrogen, D, has been used parametrlcally to show 
its effect. The value's of 0.85 and 1.0 cm 2 /s correspond, respec- 
tively, to temperatures of 630 and 730 K. Here. h SJ Is the theoreti- 
cal steady state-flame height, u Is the gas velocity, and Rj is the 
inner (nozzle) radius. Axial diffusion has been lnoluded in the 
analytical solution. 



12 
11 
10 

l» 




° ° ° ° 

• o°° 

o 

° . ' * » A * 4, A . 4 * • 








2 


* 

'* . ■ ; 8 S 8 b s 8 a a 

.So-" 8 

I ° IM Qf 1. 1 

SwmhI Onm 


JjaWI 


•Kill 

■Mtat(o>) 


iiuMil 


• 
A 

a 
o 

A 

a 


1 

t.ma 

1 


i.i 
l.a 
•.I 
l.l 
i.a 
•.■ 













6.0 



4.0 - 



<u a* 



2.0 





• • • • 

• • • • 




apOj 3 ctn cooctpoacnc*P* 3 % c * 


0° 


A 




d* 




8 




■ S« o°. 




k § 




<£ * 




- *f 

j 1 — 




o NOUML-GRAVITY KMID0N 


• LOW-CRAVHY IGNITION 

III! 



1.2 1.6 2.0 2A 23 3J 3.6 

T»»» (Me) "• 



-.4 0.0 0.4 0.8 1.2 

TIME. SECONDS 



1.8 



2.0 



2.4 



Fig. 3. Flame heights for propane following 
Ignition in mlcrogravlty; reproduced 
from [11]. 



Fig. 4. Comparison between the normal-gravity 
[7] and mlcrogravlty ignition of a 
methane flame with nozzle radius • 
0.082S cm and fuel-flow rate - 3.0 
cm 3 /sec: reproduced from [11]. 



flow in the analytical model. In Fig. 6, the effects of diffusion coefficient 
and axial diffusion are shown for a transient methane-air flame. It can be 
seen that the choice of diffusion coefficient has a significant effect on both 
approach toward steady state and flame height. However, the flame maximum 
radius is hardly affected by diffusion coefficient. In addition, the effect 
of axial diffusion is more pronounced at larger diffusion coefficients, which 
also cause a faster approach toward steady state. 

Figure 7 shows the effect of tube diameter-ratio on the theoretical 
flame behavior. As can be seen, the smaller the outer radius (i.e., the 






10 


" 






> 


\ 







-\ 




7 


• Y, 




« 


■ \ 




s 








4 








) 






P • 5.0 a*/. 


t 


\o.i 1 


• O.i caZ/t 


0.1 1-4 






, "M 


3- 




i i 


V i i 




1 




r, ca 



Fig. 3. Theoretical flame shape as a function 
of tlae for the solution with axial 
diffusion. The diffusion coefficient 
(CH 4 In N 2 ) Is 1.0 ca 2 /s (corresponding 
to 750 K). Methane and air have a velo- 
city of 10 ca/s. The Inner and outer 
radii are 0.318 and 30 ca. respectively. 
Coapare the steady-state solution with 
those of Fig. J. 



Fig. 8. 



Coaparlsona between the theoretical 
shapes of aicrogravity flaaes both with 
(•olid lines) and without (dotted lines) 
axial diffusion. Flaae shapes at 0.1 sec 

*!! d J ?L ,tMdy ,tate are ,hom - The effect 
of diffusion coefficient (CH 4 in N 2 ) is 
illustrated using D - 0.5 and 3.0 ca 2 /s 
(corresponding to T - 300 and 1800 K 
respectively). Here, aethane and air have 
a ve ocity of 10 ca/s. The inner and outer 
radii are 0.318 and 30 ca. respectively. 



smaller the amount of available air), the taller the flame. Also, beyond a 
certain outer radius, the flame height becomes independent of the outer 
geometry In Fig. 8, the effect of gas velocity on flame development is 
presented. The results are in agreement with the predictions from the 
classical solution that the steady-state flame height varies linearly with the 
gas velocity (i.e., volume flow rate). However, the present model shows that 
the same is true at any instant of time during flame development 



* 

i 

I 

c 



V«i-» 



ku • 4.M <■ 



f^ 



"To"- 



t 1 

Ttv. s 




Fig. 7. 



Effect of outer-tube radius (R 2 ) on 
flaae height. The aodel includes axial 
diffusion. Here, nozzle radius (X.) ■ 
0.18« ca. diffusion coefficient for CH 4 
In N 2 - 1.0 c» 2 /s. and h„ Is the theo- 
retical steady-state flaae height. Both 
fuel and air have a velocity of 25.76 ca/s. 



Fig. 8. Effects of gas velocity on predicted 
flaae height with no axial diffusion. 
Here, aethane and air have the saae 
velocity; diffusion coefficient - 1.0 
ca 2 /s, Inner radius ■ 0.318 ca. outer 
radius • 30 ca. and h M Is the predicted 
steady-state flaae height. 



8 



5. Conclusions 

Analytical solutions are obtained, to the first approximation, for the 
development of diffusion-flame shape and height with time, following igoi£ion 
in a zero-gravity environment. The solution reduces smoothly to the classical 
solution of laminar diffusion flames while, independently, axial diffusion 
becomes negligible, and steady state is approached. Quantitative agreement is 
obtained for the results of experiments conducted in quiescent air during 2 2 
seconds of microgravity , when the diffusion coefficient is chosen appro- 
priately. The model can predict the results obtained from the (steady state) 
classical solution, and extends these observations to arbitrary times during 
flame development. Times of the order of 2-5 seconds are predicted for 
approach toward steady state. Work is underway to apply the model to fuels 
other than methane, compare the results with those of the experiments 
conducted in quiescent air while igniting in microgravity [11-13], study the 
effect of secondary air flow on microgravity flames using the on-going 
experimental results [15], and extend the transient model to incorporate the 
effects of arbitrary diffusion coefficients and inlet velocities for the fuel 
and oxidizer streams. 

References 

1. Penner, S.S. and Sherman, S., J. Chem. Phva. 15, 569-574 (1947). 

2. Burke, S.P. and Schumann, T.E.W., Ind. Eng. Chem. 20. 998-1004 (1928) 

3. Penner. S.S., Bahadori, H.Y. . and Kennedy. E.M., Progress In Astronautics 
and Aeronautics: Vo l. 95: Dynamics of Flames and Reactive Systems . (J.R 
Bowen, N. Manson, A.K. Oppenheim, and R.I. Soloukhin, Eds.), American 
Institute of Aeronautics and Astronautics, New York, 261-292 (1984). 

4. Bahadori. M.Y., Li. C.P.. and Penner. S.S.. Progress in Astronautics and 
Aeronautics: Vol. 1 05. Part I: Flames and Configurations . (J.R. Bowen. 
J.C. Leyer, and R.I. Soloukhin. Eds.). American Institute of Aeronautics 
and Astronautics. New York, 192-207 (1986). 

5. Li, C.P.. Wiesenhahn. D., and Penner. S.S.. Combust. Plane . 65. 215-225 
(1986). 

6. Cochran, T.H. and Masica. W.J. , "Effects of Gravity on Laminar Gas Jet 
Diffusion Flames." NASA TN D-5872 (1970). 

7. Cochran, T.H.. "Experimental Investigation of Laminar Gas Jet Diffusion 
Flames in Zero Gravity." NASA TN D-6523 (1972). 

8. Haggard. J.B. , Jr. and Cochran. T.H.. Combust. Sci . Tech. 5, 291-298 
(1972): see also "Hydrogen and Hydrocarbon Flames in a Weightless 
Environment." NASA TN D-7165 (1973). 

9. Haggard, J.B.. Jr.. "Forced and Natural Convection in Laminar- Jet 
Diffusion Flames." NASA TP 1841 (1981). 

10. Edelman, R.B.. Fortune. O.F. . Weilerstein. G., Cochran, T.H.. and Haggard, 
JB > Jr - Fourteent h Symposium (International) on Combustion . The 
Combustion Institute. Pittsburgh. PA. 399-412 (1973). 

11. Bahadori, M.Y. , Edelman. R.B.. Stocker. D.P., and Olson. S.L.. AIAA J 
28, 236-244 (1990). "' 

12. Bahadori. M.Y. and Stocker. D.P., "Oxygen-Concentration Effects on Micro- 
gravity Laminar Methane and Propane Diffusion Flames," Fall 1989 Eastern 
Section Meeting of The Combustion Institute, Albany, NY (1989). 

13. Bahadori, M.Y. . Stocker. D.P., and Edelman. R.B.. "Effects of Pressure on 
Microgravity Hydrocarbon Diffusion Flames." Paper AIAA 90-0651. AIAA 28th 
Aerospace Sciences Meeting. Reno, NV (1990). 

14. Edelman, R.B. and Bahadori, M.Y. . Acta Astronaut lea . 13, 681-688 (1986). 

15. Stocker, D.P., "Size and Shape of Laminar Burke-Schumann Diffusion Flames 
in Microgravity." Spring 1990 Central States Meeting of The Combustion 
Institute. Cincinnati, OH (1990). 

16. Williams, F.A. . Combustion Theory . 2nd Ed., The Benjamln/Cummings 
Publishing Company, Inc., Melno Park. CA (1985). 



Section 11 

'EFFECTS OF PRESSURE ON MICROGRAVITY HYDROCARBON 

DIFFUSION FLAMES" 



M. Y. Bahadori, D.P. Stocker, and R. B. Edelman 



Paper AIAA-90-0651, AIAA 28th Aerospace Sciences Meeting, Reno, 

Nevada, January 1990 



&A 



AIAA 



[AIAA-90-0 651 

Effects of Pressure on Microgravity 
Hydrocarbon Diffusion Flames 

M. Yousef Bahadorj 

Science Applications International Corporation 

Torrance, California 

Dennis P. Stocker 

NASA Lewis Research Center 

Cleveland, Ohio 

Raymond B. Edelman 

Rockwell International Corporation 

Canoga Park, California 



28th Aerospace Sciences Meeting 

January 8-11, 1990/Reno, Nevada 



For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics 
370 L'Enfant Promenade, S.W., Washington, D.C. 20024 



C- 



-7. 



EFFECTS OF PRESSURE ON MICROGRAVITY 
HYDROCARBON DIFFUSION FLAMES* 

M. Yousef Bahadori + 
Science Applications International Corporation 
Torrance, California 

Dennis P. Stocked 
NASA Lewis Research Center 
Cleveland, Ohio 

and 

Raymond B. Edelman' 

Rockwell International Corporation, Rocketdyne Division 

Canoga Park, California 



ABSTRACT 

Laminar diffusion flames of propane burning in 
quiescent air (21*0 2 -79%N 2 ) are studied under both 
normal -gravity and microgravity conditions at 
pressures of 0.5, 1.0 and 1.5 atm. The effects of 
three fuel mass-flow rates (corresponding to 0.5, 
1.0 and 1.5 cnrysec at standard conditions of 1 atm 
and 294 K) have been investigated for each 
pressure. A circular nozzle of 0.15-cm diameter 
was used for these experiments, which were 
conducted in the 2.2-Second Drop Tower of NASA 
Lewis Research Center. The microgravity tests have 
shown enhanced sooting, tip-opening and soot 
escape, combined with changes 1n color, size, and 
luminosity of the flames. The corresponding 
normal -gravity flames do not show these 
characteristics. The observed differences are 
attributed to the effects of buoyancy which, once 
removed, result In increased residence time, 
dominance of diffusive processes, and extensive 
radiative loss combined with cooler overall flame 
temperatures. The paper presents the results In 
terms of the sooting behavior, Reynolds-number 
effects, pressure effects, tip-opening, and overall 
flame characteristics for both normal-gravity and 
microgravity flames. 



•paper AIAA S>6-6«1, presented at the AIAA 28th 
Aerospace Sciences Meeting, Reno, Nevada, January 
8-11, 1990. 

"•"Senior Scientist; Member AIAA. Author to whom 
correspondence should be sent to the following 
address: NASA Lewis Research Center, MS 500-217, 
Cleveland, Ohio 44135. 



t 



NASA Project Scientist. 



'Project Scientist; Member AIAA. 



This paper is declared i work of the U.S. Government and 
u not subject to copyright protection in the United Sum. 



L — Introduction 

Laminar diffusion flames of hydrogen and 
hydrocarbons burning In quiescent air'"* or with 
secondary air flows' have been studied 1n the 2.2- 
Second Drop Tower of NASA Lewis Research Center. 
In these studies, the flames were Ignited In normal 
gravity, and after attaining steady state, were 
subjected to microgravity. It was shown that 
microgravity gas jet diffusion flames are wider, 
taller and sootier than their normal -gravity 
counterparts, and are flicker-free. A review of 
these works 1s presented elsewhere.* Recent 
studies 10 - 11 have shown that by Igniting the flame 
in microgravity, improving the photographic 
technique, and reducing the filming rate, flames 
which were reported as extinguished,* reached a 
near-steady state during the 2.2 seconds of 
microgravity. Ignition in microgravity has the 
advantages of eliminating the residual buoyancy- 
induced momentum (from the pre-drop normal -gravity 
period) and not subjecting the flame to a step 
change 1n gravity. 

The present Investigation of the effects of 
pressure on microgravity laminar diffusion flames 
seeks to gain a better understanding of sooting, 
extinction, enhanced burning, and In general, the 
physico-chemical phenomena Involved in the behavior 
and characteristics of these flames. This 1s 
Important not only from the fundamental standpoint, 
but with potential applications to spacecraft fire 
safety and control of combustion in low-gravity 
environments. 



2. Experimental 



Laminar diffusion flames of propane were 
studied In quiescent oxidizing environments In the 
2. 2 -Second Drop Tower of NASA Lewis Research 
Center. For each microgravity flame, the 
corresponding normal -gravity flame was also 
studied. Details of the experimental hardware and 
procedure can be found elsewhere. 10, A circular 
nozzle (0.15 cm In diameter) was used to burn 
propane in air (21* oxygen, 79% nitrogen). The 






.04 m 5 . T B ? " 5 M, 1 S ,„d 1.5 atm. For each 
pressures of O.s, *- u * f , rales we re used 

properties and "»"'« "f" tes K ". th.t the cold- 

pressure. 

1.0 .nil 1.5 it», r,s 5«": , ;l' fi »*. The films 
£ ,"?he K"S 55." ^"ere f.rc.-pr.c.».< 1 

Sis s : sk""* 2- .'•««*■ 




3 Rgsulti 

Tlb1e 1 shows the data obtained 
aicrogravity *"« "* £ V. nonil -gravity 
tnree fuel « ss ; f1 ^hle 2 with the flame height 
data, are shown in Tabl *; J l imJ B inimum heights 
being the average of maximum » tne rlt ios of 
(1 «*, flicker zone), ***'•*,* 1on , burn-off and 
Visible soot length due to J*s sion ; ^ ^ 
SO ot quenching, as ^11 be «en i.^ ^ of 
height for "^^' ^uy . neignt , nd 



Table 1 



Mi 



crogravity data for propane flames 




•Reynolds number 



is based on the nozzle radius 



rimilarlv In Tables 2 and 3. 
and cold-fuel properties; s1«""iy 



Table 2 



Normal-gravity data for propane fl«» 



Flame 
no. 



Re 



P, 

atm 



Volume 
flow, 
cc/s 



Max 
height, 
cm 



name H1n. flame Ave flame 
height, height, 
cm 



cm 



Max. 

dia. 

cm 



Blue base 

height, 

cm 




Table 3 



K'S.i.-JC,.- «-.« 




•Here, 



-Si. 1 ' X , fl uE P S! fc ; - - 1. «* n ame 

corr es p ond , respectively ?' MX * °-9« a " d 1-9 
«x1«um. ■icrog°,"{! ,, *; d to Soot length, 9 
3 avit >' and normal. gravity 



both 2SB JrlJjg JJ "«■ shapes and color, , n 

different pressures in ?Tm-! d the thr " 
range is shown £ this "w^"'^ fl1cker 
g^vity flaws. 9 e for tne normal- 

w"»^!r. r ;«/n i i , ;; , M 2 s^™ *» ., Birili for 

the average fl ame heioh? w P r «*ure Increases 
exception for nil JtVo IE?" , 1 (w,th »" 
decreases, height of the m, "?*■ f1 *»« radius 
paction »ne becomes thinner n? decr »"s, the 
increases, and the fiJ£ k * fl1c ker distance 
However, ■IcrSgJJ^S*??^**' brighter. 
F, 9- 1); they gtn.ran» K!! d ° not fl1 "er (see 
escaping thriugh t h " tf D ha a V !. t p ;?. ti P* «Uh loot 
normal -gravity flSs Jj« t!m !? r than avera 9 e 
than their normal Jrav ??' c untf J , \" UCh *«*" 
di™* r and sootier In adaMf?^***' and »re 
increases, microgram? £ ° : " th « Pressure 
orange colors neir th. t lD^h-S°!Ti ■ 0r « red lnd 
^creases, ,„d their ^i^^^^^" ]^h 

Klcrogravlty flam* h*4„h* 
function of fuel R.vm"*H. 9h J s ar * Pitted as a 
Afferent PrlSJrl^JfLJT?: *» «■• l! for 
as a function of pr.jsur2 far wf5?" the JM » data 
"«">bers. n app,irs""2 nL'"?''"!" 1 ^ynolds 
flame height coTr^,?H«?, i 2 ind 3 tn " th. 
number but reachls 1 ^^^ > * "th. Reynolds 
»nd 1.5 atm. The flia, hi „h \ " f™ bttwttn 0.5 
"icrogravlty tests waTh,fi2 1n . a11 of these 
red soot changes coTor fnt'V?, tht 7ocit1 »n "here 
ocal nMt .* t1nc C t ° ° n r ;,'•*•" rrt, Indicting'" 

Experimental studl.shJJ.fc "'" ,n Fl 9- ')• 

burn near 1300K Th. *I! ? ho ?" th «t soot cases t» 

iffi J° r °y?^ c - "spin- to'" „;:f ^"""n 
Nation point o^St* ^ * ^"^ 




fuel Reynolds niilr -12% / ?' 15 cm ind 
nozzle radius) ThI 7 < 6ased on the 
observed are „ fL?/ ar<0 B us , Co1 °rs 
BB (brfgh? ME, f0] J°E ; » ^lue), 
blue). OP (dark BlnM (d J rk ' 0B (darl( 
(orange) „ ''^ (d "U r ed). 

is al so Seated J or 22.?' fl1Cker (f) 
flames; scale: L__^ ""j™^- gravity 



jSftS* 




Fig. 2 M1crogr.»1ty "•■•/J^iJSswr.s. 

M9 Reynolds number ^.J^'J^lnts. Note 
The lines connect the data poin 
that the flame height is ^f e 
atm and smallest at 1.0 atm. 
with Fig. 3. 




s* ?« wars*;- <• *'""-• 

compare with Fig. Z- 



.,n.„ this determination 

and ° r // 9e it MhS'dS tions .night add some 
is difficult and other oeiinu criteria are 

additional effects *!7^ flame height fof Qpen . 
£ ing looked at to define a flame ^ ^ effecU 
tip microgravity flakes »" d ^° s ° he actU a1 flame 
focalron" 9 TniVsubject 5?1 be discussed in more 
detail in the next section. 

for the (flickering) normal gr»v^ ehts ^ 
m inimum and average normal gra y Uy fUmes 

Table 2) behave similar to t re Thlt ,s. 

in relation to the effec ° T " the rJnge .5 atm 
again there exists* PT«"^ 1mum and average 

<! <1.5 atm at wh JJ th « J^", BinlBum . This is 
normal -gravity heights »tta i n . n th 

consistent with the o*£JJ that the maximum height 
microgravity flames. Note "" lnost the 

of the normal -gravity flame s »J"° (see Flg . 

same behavior ^\"* p ? Y ? e P t nat <" "9- «• the 
\\ It is interesting to note in. 
fiame of ^west Reynolds numb .r for 0.5 ^ 
rthlleTn^^irmicrVavity (Table 1). 

F i gu r, 5 shows the .j-t.^J.TJoS a d t 6f the ed 
in this\ork »• tN^-Jth of Jh . «J pressu Jnd 

flame tip). wh1c ^l" Cr indicating enhanced soot 
fuel Reynolds number, < nd ; c ^ ue % ase he ight (i.e.. 
formation. In Fig. 6, tne d . Qr t0 

decrease in length witn ui „i C rogravity 

rate. Finally. Fig. 7 g^™",^^ Note that 
flame-tip diameter vs. JJJJ{^ nt , n the beginning 
in F1g - 7 ' tt at 5 atm no attempt has been made 
USSnffS! ?nree 5 d.ta'points. 



M/iwl'» *»*« r 



1, P1»i' ss i° n 







Recent normal -graviUStw ^ increase 

flames" h» v « ,hc,m .« f V taot formation and a 
results 1n an Increase " * 00 * d °™ on . This agrees 
Xcre.se in soot remova by ox Wttj^ nontt1 . gr ,vity 

with the P rese " t < ^!'I„ Increases due to more 
flames, since brightness Incre noma1 . grlvlt y 
soot. »nd "ence, more «■ mi ° e$j 

flames. In • 1cr0 * r .i!„f , the yello* region 
ncreases. the length of th y« re<J , ns 

ISSS^-^OT^. i). ^" t,n9 thJt 

M n soot has been formed. 

t^ «~i« ^r.^ron^ u, b iro°ff y - S 

soot formation, c"^ 1 * 1 ; ,", ".jsure increases, 
pressure-dependent, to the p flQw ye1ocUy 

residence time **«"« pressure increase 
greases. I""* - "! / 1 *,, collision frequency of 
Suits in an ^{"J" J^S In coagulation, 
the particles, and hence, v* 



SPSS55W 



«<»: a 1 eh , !,°" 1 f is;s , dep r ds - «» PiPtle] . 

bum-off. B ut part c u h " 9 fictor in Partfc e 
dependent, and aTt e pr e^ Pe ° f f «» oxygT ,CU - 
oxygen partial pressure ?«i* ,ncr e»ses, the 
there exist different 'or^r"" 5 - Therefore 
competitive rates o?f ue P , re "" r f -pendent, ' 

Station "* "urn off Py w r h ° ^ h S,S lnd » r "d* 
importance of the r»i>*< ' * n,cn suggest the 

burn-off conditions r„ e ,!2P!5 ted *>r typ c \i 

Jn both non..1-gra v i ty ,J dd i t,0 n. »"« other col or. 

"> hydrocarbon «»■•. P'rticular species ...... » 

to«rd «,,„ ,J, c { 1 •«• »PM«r jr,',„ y,^ 

particle is yellow, f.J {1^' the b W>9 
: 58-ua wavelength nin/ T fb] * r »d<atfon of 

PressJr, ,"^ "*" '" Table 3, the lower th. 

— n „ »oT^re^ c ^^£r- 




^ncti „ P o ; f V J res f s U u e e Re ^°i ds »-*.r as a 
connect the data po-nts SOhdl,nes 




fig. 6 



M 

Blue-base height vs f„»i « 
at different pressur .* f Reyno,ds num ber 
"««es (open S e nd f » r "or»al-g r a V f ty 
and microgravity f >* , d f Shed ,,n «> 
"d solid lines? uZlJ "" s ^ b ° Is 
gravity and Jc ograv tTdV* 6 ?on*,. 
only the n.icrogra V i! ys ^ 1 ta . Co,ncid e, 

^e lines connect tS^aTXints^- 




KTr'Jf d^enV" ' Ue1 «^-» 
gravity f l„L | , ?L P r , " ur « f ° r "fcro- 

potnt, y S g f S i« k eo .TfS B » the data 

the smallest R*vn«iwf of t 1nform ation on 
the tip opens nmbtr for » hl 'c» 

<2& SSSA 



Since normal -gravity flames flicker, the soot 
length was not tabulated for these flames. The 
ratio of microgravity to (average) normal -gravity 
flame height increases with decrease in pressure, 
and is fairly uniform over the range of Reynolds 

numbers studied. This shows that as pressure is 
raised, microgravity and normal -gravity flame 
heights may become equal. However, the ratio of 
microgravity to normal -gravity maximum flame 
diameter does not yield sufficiently uniform 
behavior to draw a similar conclusion. 



5. CONCLUSIONS 



Laminar diffusion flames of propane burning in 
quiescent air have been studied under both normal - 
gravity and microgravity conditions at 0.5, 1.0 and 
1.5 atm. For each pressure, three fuel -flow rates 
are used. The results show the following 
conclusions: 

1. Normal -gravity flames flicker, and the 
flickering range is dampened as pressure 
decreases. Microgravity flames are steady 
and flicker-free, and are taller than their 
normal -gravity counterparts. The effect 
becomes more pronounced as the pressure 
decreases (i.e., jet velocity increases). 

2. Microgravity propane flames generally have 
open tips with soot quenching and escape at 
the tip, and beyond a certain Reynolds 
number, the flame-tip diameter becomes 
independent of pressure. At each pressure, 
there may be a well-defined Reynolds number 
below which the microgravity flame can not 
have an open tip, corresponding to a non- 
flickering normal -gravity flame. Also, the 
smaller the normal -gravity flicker range, the 
smaller the tip diameter in microgravity. 

3. The higher the pressure, the shorter the soot 
inception height, resulting 1n a faster 
approach toward sooting in both normal - 
gravity and microgravity flames. This causes 
more extensive sooting in microgravity with a 
larger soot-quench zone. Previous normal - 
gravity studies have shown that sooting is 
enhanced and burn-off is depressed by 
increase in pressure. Microgravity flames 
follow the same trend. 

4. At low pressures, ltss-eff1dent burning and 
a broader flame zone exist due to the 
kinetics effects. At high pressures, more 
soot and longer soot-burning zones make the 
task of identifying the gas-phase flame front 
difficult. For the range of pressures 
studied, there exists a particular 
intermediate pressure at which both the 
microgravity and normal -gravity flame heights 
are minimized. 

The type of data obtained in this study 
indicates that more testing Is required at 
intermediate, lower, and higher pressures, In 
addition to a wider range of Reynolds numbers to 
more fully characterize the effects of pressure on 
microgravity diffusion flames. Oata on soot 
formation and disposition need, In their own right. 



a separate study. Gathering these data will 
represent a unique source of Information for 
unraveling the complex processes Involved in 
sooting flames. Furthermore, additional analyses 
(beside the on-going theoretical effort) need to be 
performed for defining the test conditions. 
Although 1t 1s evident that the effects of buoyancy 
can be used beneficially to enhance the data base 
on soot formation and disposition processes, 1t is 
essential that tests involving non-soot1ng flames 
be conducted as well. Such tests are planned to 
sort out the effects of soot on the Interpretation 
of data, in addition to providing a way for 
diagnosis of both sooting and non-sooting flames 
under different environmental conditions in 
microgravity environments. 



Acknowledgements 



This work is supported by NASA Lewis Research 
Center under Contract NAS3-22822. The authors 
would like to thank Or. Howard Ross of NASA LeRC 
for many helpful discussions and suggestions. 

References 

1. Cochran. T.H. and Maslca. M.J.. 'Effects of 
Gravity on lialnar fits Jet Diffusion Flim," 
NASA TN D-S872, 1970. 

2. Cochrtn. T.H. and Maslca. H.J.. Thirteenth Sympo- 
sium (international) on Combustion , The Combustion 
Institute, Pittsburgh, M. 1971, pp. 821-829. 

3. Haggard, J. 8. Jr. and Cochran, T.H., Combustion 
Science and Technology , 5, 1972, pp. M1-Z9B. 

4. Cochran, T.H. ".Experimental Investigation of 
Laminar Gas Jet Diffusion Flimes 1n Zero Gravity," 
NASA TN D-6523. 1972. 

5. Haggard, J. 8. Jr. and Cochran, T.H., 'Hydrogen 
and Hydrocarbon Diffusion Flames 1n a Weight- 
less Environment/ NASA TN D-7165. 1973. 

6. Edelaan, R.B., Fortune, O.F., Hellersteln, 6., 
Cochran, T.H., and Haggard, J. 8., Jr., Fourteenth 
Symposium (I nternational) on Combustion . The 
Combustion Institute. Pittsburgh. PA. 1973. pp. 
399-412. 

7. Edelman. R.B., Fortune. 0., and Wellersteln, G.. 
•Analytical Study of Gravity Effects on Laminar 
Diffusion Flames," NASA CR-120921. 1973. 



8. 
9. 

10. 

11. 

12. 
13. 



Haggard. J. 8. Jr., 'Forced and Natural Convection 
In Lam1nar-Jet Diffusion names,* NASA TP 1841, 1981. 

Edelaan, R.B. and Bahedorl, M.Y., Acta Asronautlca . 
13, 1986. pp. 681-688. 

Edelman. R.B., Bahadorl. H.Y., Olson, S.L.. and 
Stocker. D.P., 'Laminar Diffusion Flames Under 
Microgravity Conditions,' Paper AIAA 88-064S, AIAA 
26th Aerospace Sciences Meeting. Reno. NV, 1988. 

Bahadorl, N.T., Edelman. R.B.. Stocker. D.P.. and 
Olson. S.L., 'Ignition and Behavior of Laminar 
6a i -Jet Diffusion Flames In Microgravity. 
AIAA J. . 1990, 1« preis. 

Kent, J.H. and Uegner. H. fie.. Com butt. Scl. Tech. . 
«, 1984. pp. 245-269. 

Flower, K.L. and Bowman. C.T.. Twenty-First Sympo - 
slum (Internatio nal) on Combttstton. The combustion 
Institute. Pittsburgh. PA. pp. 1114-1124. 1986. 



ORIGINAL PAGE IS 
OF POOR QUALITY 



Section 12 

'RADIATION FROM GAS-JET DIFFUSION FLAMES IN 
MICROGRAVITY ENVIRONMENTS" 



M. Y. Bahadori, R. B. Edelman, R. G. Sotos and D. P. Stacker 



Paper AIAA-91-0719, AIAA 29th Aerospace Sciences Meeting, 
Reno, Nevada, January 1991 



G&lklJkJk 



AIAA-91-0719 



Radiation from Gas-Jet Diffusion Flames 
in Microgravity Environments 



M. Yousef Bahadori 

Science Applications International Corporation 

Torrance, California 



Raymond B. Edelman 

Rockwell International Corporation 

Canoga Park, California 



Raymond G. Sotos and Dennis P. Stocker 
NASA Lewis Research Center 
Cleveland, Ohio 



29th Aerospace Sciences Meeting 

January 7-10, 1991/Reno, Nevada 



For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics 
370 L'Enfant Promenade, S.W., Washington, D.C. 20024 



RADIATION FROM GAS-JET DIFFUSION FLAMES IN MICROGRAVITY ENVIRONMENTS* 



M. Yousef Bahadori 

Science Appl ications International Corporation 

Torrance, California 

Raymond B. Edelman 

Rockwell International Corporation, Rocketdyne Division 

Canoga Park, California 



and 

5 f 

Raymond G. Sotos and Dennis P. Stocker 

NASA Lewis Research Center 
Cleveland, Ohio 



ABSTRACT 



Introduction 



This paper presents the first demonstration 
of quantitative flame-radiation measurement in 
microgravity environments, with the objective of 
studying the influences and characteristics of 
radiative transfer on the behavior of gas-jet 
diffusion flames with possible application to 
spacecraft fire detection. Laminar diffusion 
flames of propane, burning in quiescent air at 
atmospheric pressure, are studied in the 5.18- 
Second Zero-Gravity Facility of NASA Lewis Research 
Center. Radiation from these flames is measured 
using a wide-view angle, thermopile-detector 
radiometer, and comparisons are made with normal- 
gravity flames. The results show that the 
radiation level is significantly higher in micro- 
gravity compared to normal-gravity environments due 
to larger flame size, enhanced soot formation, and 
entrapment of combustion products in the vicinity 
of the flame. These effects are the consequences 
of the removal of buoyancy which makes diffusion 
the dominant mechanism of transport. The results 
show that longer test times may be needed to reach 
steady state in microgravity environments. 



*Paper AIAA-91-0719, presented at the AIAA 29th 
Aerospace Sciences Meeting, Reno, Nevada, January 
7-10, 1991. 

+ Sen1or Scientist; Member AIAA. Author to whom 
correspondence should be sent to the following 
address: SAIC/NASA LeRC, MS 500-217, Cleveland, 
Ohio 44135. 

*Ch1ef Scientist; Member AIAA. 

s Manager, Zero-Gravity Facility. 

'NASA Project Scientist. 



This paper is declared a work of the U.S. Government and 
is not subject to copyright protection in the United States. 



The term "diffusion flame" classifies those 
types of flames 1n which the fuel and oxidizer are 
not premixed. The basic mechanisms which control 
these flames (and combustion 1n general) include 
the coupled processes of mixing, chemical kinetics 
Including soot formation and disposition, flame 
radiation, and buoyancy- Induced convection. A 
major problem Impeding our understanding of flame 
behavior is the masking effect of buoyancy on 
flames that are studied in a normal -gravity 
environment. The behavior of fires on earth 1s 
determined by the combined convectlve and diffusive 
effects, where buoyancy- Induced convection masks 
the processes of chemlcal-dlffuslonal Interaction 
that are fundamental to the understanding of 
combustion phenomena. In zero-gravity environments, 
the buoyant force is eliminated, and the remaining 
physico-chemical processes are responsible for the 
different behavior of flames observed in the micro- 
gravity experiments conducted to-date. 1 

Recent studies of microgravity laminar 
diffusion flames have shown some unique, and in 
most cases, unexpected phenomena. 2 " 7 For example, 
in order to study the effects of microgravity on 
flammab111ty and extinction of flames In the near- 
limit regions, tests were conducted under reduced- 
oxygen 2 and/or reduced-pressure 1 conditions. The 
results show that flame characteristics, color, 
luminosity, and sooting behavior are strongly 
affected 1n microgravity compared to normal 
gravity. 2 Sooting was eliminated in microgravity 
propane flames at 18X oxygen, and the flames were 
entirely blue, whereas their normal-gravity 
counterparts were yellow, luminous, and very 
similar to flames under atmospheric conditions. 
This has an Important Implication, namely, reduced 
radiative heating and reduced hazard of flame 
spreading to the surrounding combustible materials 
by reducing the oxygen concentration In micro- 
gravity environments. In addition, flammability 
limits may be quite different in microgravity and 
normal -gravity flames under Identical conditions of 
pressure and oxygen concentration. The present 
study was conducted to investigate the effects of 
radiative loss on flame behavior and characteristics. 



of ?°°* 



2. Experimental 

The experiments were conducted 
Second Zero-Gravity Facility of NASA 
Center. Figure 1 shows a schematic 
The volume of the sealed combustion 
0.087 m 3 . The tapered-tip nozzle ha 
diameter of 0.17 cm. Both normal-gr 
gravity flames of propane were studi 
oxidizing environments of 215 oxygen 
at 1 atm. The fuel volume-flow rate 
and 1.5 cm 3 /sec, corresponding to a 
nolds number of 45, 90, and 135, res 



in the 
Lewis 
of the 
chamber 
d an in 
avity a 
ed in q 
791 ni 
s were 
cold-je 
pective 



5.18- 
Research 
hardware . 
was 
iner 

nd micro- 
uiescent 
trogen 
0.5, 1.0 
t Rey- 

ly. 



The flames were ignited using a spark electrode 
located 1.0 cm above the nozzle tip and 2.0 cm away 
from the nozzle centerline, with the nozzle acting 
as the ground electrode. The ignition system was 
activated for approximately 1.0 sec for both the 
normal-gravity and microgravity flames. The micro- 
gravity flames were ignited after the release of 
the package in the evacuated drop tube. Two movie 
cameras (filming rates ' 12 and 30 frames/sec, res- 
pectively) using Kodak Ektachrome Video News films 
recorded the flame development and behavior. 

The wide-view-angle radiometer was comprised 
of a 1.5xl.5-mm, fast-response thermopile detector 
with a Calcium Fluoride (CaF 2 ) window (Oriel Corpo- 
ration, Model 7106), with a time constant of 40 
msec, field of view of 90-100°, and maximum irra- 
diance of 0.2 W/cm 2 . In addition, the experiment 
chamber was equipped with a CaF 2 window to allow 
maximum transmission of the infrared radiation. 
The window had a usable transmission range of 
0.2-9.0 urn, with its absorptivity falling off above 
1.0 urn to about 50X at 10.0 um; its approximate 
transmission in the linear range was 94X. The 
inside of the chamber was painted black to avoid 
reflection from the wall. 



The 
body sou 
with an 
ter (Kei 
measure 
tures of 
the nozz 
the blac 
28.6 cm, 
was 3.84 
meter of 
Fluoride 
iris/sen 
flame co 



radiometer was calibrated using a black- 
rce (Barnes Engineering, Model 11-210), 
opening of 2.54 cm in diameter. A nanome- 
thly Instrument, Model 181) was used to 
the radiometer response to various tempera- 
the black-body source. The distance from 
le centerline (and for calibration, from 
k-body source) to the radiometer sensor was 
and from the shutter iris to the sensor 
cm. The Iris on the shutter had a d1a- 
2.54 cm. The diameter of the Calcium 
window was 4.45 cm. The nozzle/window/ 
isor arrangement was such that the entire 
uld be seen by the sensor. 



MS] MTU 




min cncu 



vmi aimoa 



nu. mmi 



Fig. 1 Experiment package for the 5.18-second 
microgravity tests. 



Results 



Figure 2 shows the flame shapes and colors for 
the three flow rates under both normal-gravity and 
microgravity conditions. The normal-gravity flames 
were flickering with a typical flicker frequency of 
laminar diffusion flames 1n the absence of air flow 
However, the microgravity flames developed conti- 
nuously toward steady state during the 5 seconds of 
available time, and the shapes presented 1n Fig 2 
are for the last few fractions of a second. 

Figure 3 shows the flame radiance as a function 
of time. The radiometer registers the data in terms 
of mV. Using the calibration chart and converting 
the mV to the equivalent black-body temperature of 
the calibration source, radiance (oTg/n) can be 
obtained, where a 1s the Stefan-Boltzmann constant, 
and Tb is the black-body temperature. 

In Table 1. the data obtained for both normal- 
gravity and microgravity flames are presented in 
terms of flame height, maximum flame diameter, 
radiance, and radiance/flame height. Data interpre- 
tation will be discussed 1n the next section The 
radiance data are the average of the values during 
the last second of the experiment. 

Figure 4 shows the radiance as a function of 
fuel-volume flow rate for both normal-gravity and 
microgravity flames. 



Volume 
flow 
rate, 

cm'/sec 



0.5 
1.0 
1.5 



Table 1 Normal -gravity and microgravity data for propane 
flames burning In quiescent air at 1.0 atmosphere 



Normal -Gravity Flames 

Ave. flame Max. flame Flame 

-adiani 
U/cm 2 -sr 



height, diameter, radiance, ? ad1ance 
cm rm u/,-«,2 ,_ Ave. Ht. 



cm 



2.1 
4.1 
4.9 



•v-0.6 
V).7 
-vO.7 



0.18 
0.34 
0.54 



0.09 
0.08 
0.11 



3.5 

5.9 

10.3 



Microgravity Flames 



Flame Max. flame Flame 
height, diameter, radiance, Ra , d ] a " ce 
cm cm W/c« , -sr H e 1 9" 



2.3 
2.4 
2.3 



1.29 
2.31 
4.68 



0.37 
0.39 
0.45 



ORIGINAL PAGE IS 
OF POOR QUALITY 



0-g 



3, 



0.5 cm /sec 




o-g 

3 
1.0 cm /sec 



o-g 

1.5 cm /sec 





1-9 

0.5 cm /sec 



B/p— y 

DB/P-^ 1 



1-9 

1.0 cm /sec 



1-9 

3 
1.5 cm /sec 





Fiq. 2 Normal-gravity and microgravlty flames of 
propane, at different volume flow rates, 
burning 1n quiescent air at 1.0 atm. The 
various colors observed are as follows: 
B (blue), DB (dark blue), DR (dull red), 
(orange). P (pink). R (red), W (white), 
and Y (yellow). The range of flicker, f. 
1s shown for the normal -gravity flames; 
scale: *— « ■ 1 cm. 



— 12 














CM 

■ a 


g ■ 0.5 cn^/sec 








u 8 




















0-9 




s * 

* 


._^c= 






4- 





12 3 4 5 6 

TIM (SK) 



12 



g 4 

















g - 1.0 oP/stc 


























0-9 


1-9 










/ - 


X- 








"-■ — «— « 





1 2 3 4 5 6 

T<at (stc) 



« 12 

1. 












Si ft 




Q ■ 1.5 c»V*tc 






8 








r 






V 


r* 






/ 






c 4 

•a 








1-9 


S 


h~ 








^J- 


^ 



TIM (Stc) 



Fig. 3 Measured radiance vs. time for both 

normal-gravity and mlcrogravlty flames of 
propane burning In air at 1 atm. 




o.o 



0.5 l.o 1.5 

Voliat FIom ItU |o 3 /«»c) 



Fig 



I Radiance as a function of fuel volume- 
flow rate for propane flames burning in 
air at 1 atm. The radiance data are the 
average values between 4.0 sec and 5.0 
sec after Ignition. 



£$£ *&" 



4. Discussion 

The flame shapes presented in Fig. 2 show that 
microgravity flames are taller, wider, and sootier 
than their normal -gravity counterparts. The orange- 
reddish color of these flames is an indication of 
cooler temperatures than is expected for typical 
burn-off conditions. The yellow/orange/red/dull red 
regions in the flame show that the temperature of 
the soot decreases in microgravity flames, leading 
to local extinction at the flame tip. This tip- 
opening and underventilated-type behavior is a 
characteristic of sooty microgravity flames. 
Absence of buoyancy and dominance of diffusive pro- 
cesses result in longer residence times, enhanced 
soot formation, increased radiative loss and cooling 
by pyrolysis, which are responsible for the observed 
behavior and characteristics of microgravity flames. 
The data presented in Table 1 show that as the fuel 
flow rate increases, both microgravity and average 
normal -gravity flame heights increase, but the 
effect is more pronounced for microgravity flames. 
Although the maximum flame diameter is independent 
of fuel volume flow rate in both microgravity and 
normal gravity, the microgravity flames show a four- 
fold increase in diameter. This is due to the 
accumulation and slow transport of combustion pro- 
ducts in the vicinity of the flame in microgravity. 
The microgravity flame heights presented 1n Table 1 
are based on the location where red soot changes 
color into dull red, indicating local flame extinc- 
tion. Experimental studies have shown that soot 
quenching occurs around 1300 K. Assuming soot to be 
a black body, orange-red (M370 K), cherry red 
(M170 K) and dull red (-^970 K) clearly indicate 
that extinction occurs near the flame tip. 

The data of Fig. 3 show that radiative loss is 
constant for normal -gravi ty flames throughout the 
duration of the test. This is largely due to the 
instantaneous removal of the hot combustion products 
under the influence of buoyancy. In addition, under 
normal-gravity condition, the flame immediately 
establishes following ignition, whereas in micro- 
gravity, a continuous development of the flame is 
observed (both visually by examining the films and 
through radiation data). Figures 3 and 4 show that 
fuel flow rate has a much more pronounced effect on 
the level of radiation from microgravity flames 
compared to those in normal gravity. The radiation 
data of Fig. 4 show that radiance varies linearly 
with fuel flow rate under normal -gravity condition, 
again, due to the dominance of buoyancy, which 
limits the radiative loss. However, In microgra- 
vity, radiance quickly starts to increase as the 
amount of injected fuel increases. This Is not only 
due to the larger flame size and extensive soot for- 
mation, but also the accumulation and slow transport 
of the combustion products as mentioned before. The 
hot products of combustion (mostly CO2 and H?0) 
appear to contribute dramatically to radiative 
loss from the microgravity flames. The data of 
Table 1 and Fig. 4 show that radiative loss from the 
microgravity flames studied in this work Is 6-9 
times larger compared to the corresponding normal - 
gravity flames. In Table 1, radiance/ flame height 
1s also calculated. Since flame diameters are 
observed to be independent of flow rate for a given 
gravitational level, flame height 1s a good measure 
for obtaining radiance per unit volume of the flame. 
It can be seen that radiance/height Is ^.1 for 
normal -gravity flames and M).4 for microgravity 
flames. Whether these values change at higher flow 
rates needs to be determined. 



The data presented in this study suggest that 
radiative ignition of nearby materials may be promo- 
ted in low-gravity environments due to the increased 
radiative loss from gas-jet diffusion flames. 
Examination of the data also show that radiation may 
not reach steady state in 5 seconds of microgravity. 
This is apparently due to the continuous accumulation 
and slow transport of combustion products in the 
vicinity of the flame. Sometimes the luminous 
region does not appear to increase in size. However, 
the diffusion of non-luminous hot gases in the 
surrounding causes the volume of the hot gas to 
increase, which is reflected through the continuous 
increase in radiation (and temperature of the field). 
This has a direct impact on the question of approach 
toward steady state in microgravity flames, indica- 
ting that the reduced-gravity earthbound facilities 
may not provide enough time to obtain the true 
steady-state data. Previous studies 1 '*' 7 have shown 
that transient effects become important for micro- 
gravity laminar diffusion flames. The data obtained 
in the present study seem to substantiate these 
observations. 

5. Conclusions 

Quantitative measurement of flame radiation has 
been obtained for the first time in a microgravity 
study. Laminar diffusion flames of propane, burning 
in quiescent air at atmospheric pressure, have been 
studied for flow rates of 0.5-1.5 cm'/sec. The 
results show that flame radiation is 6-9 times 
larger in microgravity compared to normal gravity. 
Enhanced soot formation, larger flame size, and 
entrapment and slow transport of the hot combustion 
products in the vicinity of the microgravity flame 
are the contributing factors. The normal -gravity 
radiation data quickly reach steady state, whereas 
Hie corresponding microgravity data are 1n a 
transient state throughout the duration of the 
experiment. The results show that longer test times 
are needed in microgravity to reach steady state. 

Acknowledgements - This work is supported by NASA 
Lewis Research Center under Contract NAS3-22822. 
The authors would like to thank Paul Tavernelli and 
John Kolacz of NASA LeRC, Engineering Directorate, 
for design of the data acquision system and 
calibration of the radiometer. 



Reference! 

'Edelaan, R.B., ind Bahadorl, M.r.. AcU Astronautics . 13. 
pp. (81-688, 1986; see also references cited tn this aork. 

'Bahadorl, M.T.. and Stocktr, O.P., 'Oiygen-Concentratlon 
Effect! on Microgravity Laminar Methane and Propane Dlffuilon 
Flames,' 1989 Eastern State! Meeting of The Coabuitlon 
Institute. Albany, New fork. October/November 1989. 

"Bahadorl. M.r.. Stocker, D.P.. and Edelaan, «.«., 'Effects of 
Pressure on Microgravity Hydrocarbon Diffusion Flaacs.* Paper 
A1AA-9O-0651, AIAA Z8th Aerospace Meeting. Reno, NV. Jan. 1990. 

•Bahadorl, M.T.. Edelaan. R.8., Stocker. D.P.. and Olson, S.L.. 
AIAA J. . 28, PP.Z36-M4. 1990. 

"Stocker, D.P., "SIm and Shape of Laminar Burke- Schumann 
Diffusion flaaes In Microgravity," 1990 Central States Meeting 
or The Coabuitlon Institute, Cincinnati, Ohio, Hay 1990. 

•Bahadorl, M.r.. 'An Analytical Solution for Transient, Cyllnd- 
rlcally Symmetric Laalnar Diffusion Flaaes In the Absence of 
Buoyancy." 1990 Central States Meeting of The Combustion 
Institute.' Cincinnati, Ohio, Nay 1*90. 

'Bahadorl, N.T., Edelaan, R.B., Sotos. R.S.. and Stocker. D.P.. 
"Measurement of Temperature In Microgravity Laalnar Diffusion 
Flames.' 1990 Eastern States Meeting of The Combustion 
Institute. Orlando. Florida. December 1990. 



ORIGINAL PAGE IS 
OF POOR QUALITY 



Section 13 

"EFFECTS OF OXYGEN CONCENTRATION ON RADIATIVE LOSS FROM 
NORMAL-GRAVITY AND MICROGRAVITY METHANE DIFFUSION FLAMES" 



M. Y. Bahadori, R. B. Edelman, D. P. Stacker, R. G. Sotos, and D. F. Vaughan 



Paper AIAA-92-0243, AIAA 30th Aerospace Sciences Meeting, 
Reno, Nevada, January 1992 



gi 



AMAA 



AIAA-92-0243 

Effects of Oxygen Concentration on Radiative 
Loss from Normal -Gravity and Microgravity 
Methane Diffusion Flames 



M. Yousef Bahadorl 

Science Applications International Corporation 

Torrance, California 

Raymond B. Edelman 

Rockwell International Corporation 

Canoga Park, California 

Dennis P. Stocker, Raymond G. Sotos, and David F. Vaughan 
NASA Lewis Research Center 
Cleveland, Ohio 

30th Aerospace Sciences 
Meeting & Exhibit 

January 6-9, 1992 / Reno, NV 



For permission to copy or republish, contact the American Institute o( Aeronautics and Astronautics 
370 L Enfant Promenade, S.W., Washington, D.C. 20024 



EFFECTS OF OXYGEN (XWCENTRATION 

ON RADIATIVE LOSS FROM NORMAL-GRAVITY AND 

MICROGRAVrrY METHANE DIFFUSION FLAMES* 

M. Yousef Bahadorit 

Science Applications International Corporation 

Torrance, California 

Raymond B. Edelmanf 

Rockwell International Corporation, Rocketdyne Division 

Canoga Park, California 



and 



Dennis P. Stocker,§ Raymond G. Sotos.t and David F. Vaughan* 

NASA Lewis Research Center 

Cleveland, Ohio 



Abstract 

Laminar diffusion flames of methane, 
burning in quiescent oxidizing environments 
at atmospheric pressure, have been studied 
under both normal-gravity and microgravity 
conditions. Radiation from these flames is 
measured using a wide-view-angle, 
thermopile detector radiometer. The oxidizer 
was 18%, 21%, and 30% oxygen in nitrogen. 



* Paper AIAA-92-0243, presented at the 
AIAA 30th Aerospace Sciences Meeting, 
Reno, Nevada, January 6-9, 1992. 

t Senior Scientist; Senior Member AIAA. 
Author to whom correspondence should be 
sent to the following address: SAIC/NASA 
LeRC, MS 500-217, Cleveland, Ohio 44135. 

t Chief Scientist, Member AIAA 

J Manager, Zero-Gravity Facility. 

f NASA Project Scientist 

• Baldwin-Wallace College, Berea, Ohio. 



This paper is declared a work of the U.S. 
Government and is not subject to copyright 
protection in the United States. 



The results show that radiation levels are 
significantly higher in microgravity compared 
to normal-gravity environments. This is due 
to larger flame size, enhanced soot formation, 
and accumulation and slow transport of the 
combustion products in the vicinity of the 
flame, which are caused by the removal of 
buoyancy. For the normal -gravity flames, the 
effect of oxygen concentration on radiation is 
negligible, whereas in microgravity, the flame 
radiation is strongly affected by the amount 
of oxygen in the environment This has direct 
applications to the problem of spacecraft fire 
detection and fire safety. 

L IntrorfurHnn 

The fundamental understanding of flame 
characteristics under reduced-gravity 
conditions is essential to the problem of fire 
safety in spacecraft environments. 1 The 
behavior of fires on Earth is influenced by the 
combined convective and diffusive effects, 
where buoyancy-induced convection masks 
the processes of chemical- diffusional 
interaction. Under microgravity conditions, 
the buoyant force is nearly eliminated, and 
the coupled physico-chemical processes of 
diffusion, mixing, kinetics, radiation, and 
convection become more tractable. A major 
problem impeding our understanding of flame 
behavior is this masking effect of buoyancy, 
which results in some very significant 



differences between the structure of normal- 
gravity and microgravity flames. 2 

"Diffusion" flames, as the name implies, are 
flames in which the fuel and oxidizer are 
initially "non-premixed", and the combustion 
process is governed by the diffusion of these 
reactants toward each other to form a 
reaction zone (i.e., the flame sheet). In 
"laminar" diffusion flames of gaseous fuels, as 
opposed to "turbulent" gas-jet diffusion 
flames, the flame height increases almost 
linearly with the increase in fuel-flow rate (or 
the Reynolds number). In recent studies of 
microgravity laminar diffusion flames, some 
unique and unexpected phenomena have 
been observed. 2 " 10 The results have shown 
that flame color, size, luminosity, sooting 
behavior, and approach toward steady state 
are significantly different in microgravity 
compared to normal gravity. In addition, 
these differences are exacerbated by the 
change in pressure and/or oxidizer 
composition for the microgravity flames. 31 *• 10 

Radiative characteristics of flames is the 
focus of the present study. The total radiation 
from normal-gravity and microgravity flames 
of methane have been measured. This type of 
measurement has been used extensively by 

Marksteinll'l° to study radiation from a 
variety of laminar and turbulant diffusion 
flames in normal gravity environments. 



L — Experimental 

The experiments were conducted in the 
5.18-Second Zero-Gravity Facility of NASA 
Lewis Research Center. 1 '' Figure 1 shows a 
schematic of the hardware. The volume of 
the combustion chamber was 0.087 m 3 . The 
fuel nozzle (protruding IS cm inside the 
chamber) had a tapered tip with inner radius 
of 0.0825 cm, and was positioned along the 
centerline of the cylindrical chamber. The 
nozzle was approximately 20 cm away from 
the chamber wall. The tip of the nozzle was 
approximately 50 cm away from the 
chamber top. The fuel was supplied from a 
stainless steel bottle of 75 cm 3 in volume at 
an initial pressure of 100 psig. Appropriate 
windows were mounted in the chamber wall 
for the movie cameras and the radiometer. A 
pressure transducer was used to record the 
chamber pressure during the test. 

Both normal-gravity and microgravity 



flames of methane (99.9%-pure) were studied 
in quiescent oxidizing environments of 18%, 
21% and 30% by volume of oxygen in 
nitrogen. The chamber pressure was 1.0 atm, 
and the fuel-volume flow rate was 
3.0 cm 3 /sec. 

The flames were ignited using a spark 
electrode, which was located 1.0 cm above the 
nozzle tip and 2.0 cm away from the nozzle 
centerline, with the nozzle acting as the 
ground electrode. The ignition system was 
activated for 0.5-1.0 sec for the normal 
gravity and microgravity tests. The 
microgravity flames were ignited after the 
release of the package in the drop tube. Two 
movie cameras (filming rates = 12 and 30 
frames/sec, respectively) using color Kodak 
Ektachrome Video News film recorded the 
flame development and behavior. 

The wide-view angle radiometer was 
comprised of a 1.5 x 1.5-mm, fast-response 
thermopile detector with a Calcium Fluoride 
(CaF 2 ) window (Oriel Corporation, Model 
7106), and a time constant of 40 msec, field 
of view of 90 - 100 degrees, and maximum 
irradiance of 0.2 W/cm 2 . In addition, the 
experiment chamber was equipped with a 
CaF 2 window to allow maximum transmission 
of the infrared radiation. The window had a 
usable transmission range of 0.2 - 9.0 urn, 
with its absorptivity falling off above 1.0 urn 
to about 50% at 10.0 urn; its approximate 
transmission in the linear range was 94%. 
The radiometer was calibrated using a black- 
body source (Barnes Engineering, Model 
11-210), with an opening of 2.54 cm in 




NXIC CIMM 



WW IUCTWW 



Fig. 1 Experiment package for the 

5.18-second microgravity tests. 



diameter. A nanometer (Keithly Instrument, 
Model 181) was used to measure the 
radiometer response to various temperatures 
of the black-body source. The distance from 
the nozzle centerline (and for calibration, 
from the black-body source) to the 
radiometer sensor was 28.6 cm, and from the 
shutter iris to the sensor was 3.84 cm. The 
iris on the shutter had an open diameter of 
2.54 cm. The inner diameter of the Calcium 
Fluoride window was 4.43 cm. The 
nozzle/window/iris/sensor arrangement was 
such that the field of view on a plane passing 
through the centerline of the nozzle had a 
radius of 9.55 cm. 



3. Results. 

Figure 2 shows the normal-gravity and 
microgravity flames of methane burning in 
quiescent oxidizing environments of 18%, 21%, 
and 30% by volume of oxygen in nitrogen. 
Following ignition, the normal-gravity flames 
establish quickly (in less than 0.5 sec), and 
show the typical flicker characteristics (with a 
frequency of 10 - 20 Hz) of laminar flames. 
The flicker range is also shown in Figure 2 for 
the normal-gravity flames. However, the 
microgravity flame heights continuously 
increase after ignition, and reach near-steady 
state after a few seconds following ignition. 
The microgravity flames shown in Figure 2 
are for the last instant of the drop (just prior 
to the beginning of deceleration). 

Figure 3 shows the flame radiance as a 
function of time for both normal-gravity and 
microgravity flames. Using the calibration 
chart, the data obtained from the radiometer 
reading (in mV) was converted to the 
equivalent black-body temperature (T b ) of 
the calibration source. The flame radiance 
(oTVll) in W/cm J -sr was then obtained as a 
function of time, where o is the Stefan- 
Boltzmann constant. 

Table 1 shows the data obtained in terms 
of flame height, flame maximum diameter, 
radiance, and radiance per unit projected 
flame area. This area was calculated by 
multiplying the flame height by the maximum 
flame diameter, to give a rough estimate of 
the relative values of radiance per unit flame 
area between the normal-gravity and 
microgravity flames. The radiance data are 
averages for the last second of each test The 









Fig. 2 Normal-gravity and microgravity 
flames of methane, burning in 
different oxygen-nitrogen 
environments at 1.0 atm. The 
various colors observed are as 
follows: B (blue), DB (dark blue). 
O (orange), P (pink), R (red), 
W (white), and Y (yellow). The 
range of flicker, f, is shown for the 
normal-gravity flames; 
scale: • 1 = 1 cm. 



heights for normal-gravity flames in Table 1 
are the averages of minimum and maximum 
flickering-flame height. 



— 12 



c 4 





















181 Oxygen 






0-9 












| 


1-9 












-J-- 












.L 




""^ 



-T 12 



- 12 



2 3 4 
Time (sec) 



0-9 



# 



2H Oxygen 



■vh*« 



2 3 

Time (sec) 





















30% Oxygen 






o, 












I 


1-9 














1 •■ 




^rfrL v 






i 







u 1 2 3 4 5 
Time (sec) 

Fig. 3 Measured radiance vs. time for 

both normal-gravity and 
microgravity flames of methane 
burning in different oxidizing 
environments at 1 atm. 



In Figure 4, flame radiance as a function of 
oxygen concentration is presented. Both 
normal-gravity and microgravity flames show 
an increase in radiance with increasing 
oxygen concentration. However, the effect is 
more significant for the microgravity flames. 



4.0 



3.0 



2.0 



1.0 - 



0-9 




I I I i I i i i i i I i i i i 



181 211 
Percent Oxygen 



301 



Fig. 4 



Radiance as a function of percent 
oxygen for methane flames 
burning at 1 atm. The radiance 
data are the average values 
between 4.0 sec. and 5.0 sec. after 
ignition. 



TABLB 1 NORMAL-GRAVITY AND MICROGRAVITY DATA FOR METHANE FLAMES 
BURNING IN QUIESCENT OXIDIZING ENVIRONMENTS AT 1.0 ATM 







Normal-Gravity Flames 




Microgravity Flames 


Oxyt«n 


Avf. flam* 
haiiM. 
cat 


Mai. nana 
diaaeter. 
cat 


Flaaa 
radianca. 
W/ea'-ir 


Radiance 
Flam* araa 


Flam* 
hetaM. 
ca 


Mai. flaaa 
dlaaatar. 

ca 


Flaaa 
radiaaca. 
W/ea»-tr 


ladlauct 
Flaaa araa 


III 
211 
30* 


3.2 
4.1 
3.1 




1.3 
1.2 
I.I 


0.24 
0.32 
0.3« 


0.04 
0.07 
0.11 


a.2 

7.7 
3.-4 


2.4 
2.2 
1.7 


2.63 
2.93 
3.33 


0.13 
0.17 
0.36 



4. DliKUMinn 

The flames of Figure 2 show the significant 
effects of buoyancy on size, shape and 
luminosity. The normal-gravity flames are 
generally yellow, flickering, and pencil-like, 
regardless of oxygen level in the 
environment. However, their flicker range 
decreases and their intensity appears to 
increase at higher oxygen concentrations. The 
microgravity flames, on the other hand, are 
taller, wider, and more diffuse than the 
normal-gravity flames. As the oxygen level 
increases, the flame becomes brighter due to 
higher gas-phase and particulate 
temperatures and faster soot burn-off. 

The orange/reddish color of the 
microgravity flames is an indication of cooler 
temperatures than is expected for typical 
burn-off conditions. The yellow/orange/red 
regions in the flames show that the 
temperature of the soot decreases in 
microgravity flames. Absence of buoyancy 
and dominance of diffusive processes result in 
longer residence times, enhanced soot 
formation, increased radiative loss and 
additional cooling by pyrolysis, which are, in 
some combination, responsible for the 
observed characteristics of microgravity 
flames. 

The data presented in Figure 3 show that 
radiative loss in normal-gravity flames is not 
only constant throughout the duration of the 
test, but that the level of radiation does not 
change significantly with increasing oxygen 
concentration. This is largely due to the rapid 
removal of the hot combustion products 
under the influence of buoyancy. In addition, 
the normal-gravity flames become 
established immediately following ignition, 
whereas in microgravity, the flame increases 
in height for a few seconds following ignition 
(both visually by examining the films and 
through radiation data). 

Table 1 shows that for both normal-gravity 
and microgravity flames the radiation 
decreases with reduced ambient oxygen 
concentrations. It is interesting to note that 
the radiation from the normal gravity flames 
falls within a narrow range, as can be seen in 
Figure 4. Finally, Table 1 shows a linear 
variation of flame height with change in 
oxygen concentration for both the normal- 
gravity and microgravity flames. This is due 



to the availability of oxygen in the 
environment. The data presented in Table 1 
also show that radiance per unit flame area is 
significantly different between normal- 
gravity and microgravity flames. This is 
attributed to the accumulation and slow 
transport of the hot combustion products in 
the vicinity of the microgravity flame. 



Laminar diffusion flames of methane have 
been studied under both normal-gravity and 
microgravity conditions. The quiescent 
oxidizing environment was composed of 18%, 
21%, and 30% oxygen in nitrogen at 1.0 
atmosphere. A wide-view-angle, thermopile- 
detector radiometer was used to measure the 
overall radiative loss from these flames. The 
results show that flame radiation is an order 
of magnitude higher in microgravity 
compared to normal gravity for the flames 
studied in this work. Larger size of the 
microgravity flames caused by the removal of 
buoyancy and the accumulation and slow 
transport of the hot combustion products in 
the vicinity of the flame is largely responsible 
for this observation. The normal-gravity 
radiation data reach steady state in less than 
1.0 second, whereas the microgravity data 
take up to S.O seconds to reach near-steady 
state. The normal-gravity flames do not show 
appreciable effects of oxygen concentration, 
whereas those in microgravity show 
significant increase in radiation with 
increasing oxygen concentration. 



Acknowledgment! 

This work was supported by NASA Lewis 
Research Center under Contract NAS3-22822 
with Science Applications International 
Corporation. D. F. Vaughan was supported by 
Grant NCC-3157 with Baldwin- Wallace 
College. 



References 

Sackstedcr, K. R.: Twentv-Third 
.Symposium (International^ on 
Combustion. P. 1589, The Combustion 
Institute, 1990. 

Edelman, R. B. and Bahadori, M. Y.: A£lA 
Astronautica 13, 681 (1986); for 
previous studies of microgravity 
laminar diffusion flames, see references 
cited in this work. 

Bahadori, M. Y. and Stocker, D. P.: 
Oxygen-Concentration Effects on 
Microgravity Laminar Methane and 
Propane Diffusion Flames, Eastern States 
Meeting of The Combustion Institute, 
Albany, New York, October/November 
1989. 

Bahadori, M. Y., Stocker, D. P., and 
Edelman, R. B.: Effects of Pressure on 
Microgravity Hydrocarbon Diffusion 
Flames, Paper AIAA-90-0651, AIAA 
28th Aerospace Sciences Meeting, Reno, 
Nevada, January 1990. 

Bahadori, M. Y., Edelman, R. B., 
Stocker, D. P., and Olson, S. L.: 
AIAA J. 28. 236 (1990). 

Stocker, D. P.: Size and Shape of 
Laminar Burke-Schumann Diffusion 
Flames in Microgravity, Central States 
Meeting of The Combustion Institute, 
Cincinnati. Ohio, May 1990. 

Bahadori, M. Y.: An Analytical Solution 
for Transient, Cylindrically Symmetric 
Laminar Diffusion Flames in the 
Absence of Buoyancy, Central States 
Meeting of The Combustion Institute, 
Cincinnati, Ohio, May 1990. 

Bahadori, M. Y., Edelman, R. B., 
Sotos, R. C, and Stocker, D. P.: 
Measurement of Temperature in 
Microgravity Laminar Diffusion Flames, 
Eastern States Meeting of The 
Combustion Institute, Orlando, Florida, 
December 1990. 



9. Bahadori, M. Y., Edelman, R. B.. 
Sotos, R. G.. and Stocker, D. P.: 
Radiation from Gas-Jet Diffusion Flames 
in Microgravity Environments, Paper 
AIAA-91-0719, AIAA 29th Aerospace 
Sciences Meeting, Reno, Nevada, 
January 1991. 

10. Bahadori. M. Y., Edelman, R. B., 
Sotos, R. G., Stocker, D. P., and 
Vaughan, D. F.: Effects of Oxygen 
Concentration on Radiation From 
Microgravity Laminar Propane Diffusion 
Flames, submitted for presentation at 
the Twenty-Fourth Symposium 
(International) on Combustion, 1992. 

11. Markstein. G. H.: Fifteenth Symposium 
rinteraation an on Combustion. P. 1285, 
The Combustion Institute, 1974. 

12. Markstein, G. H.: fnmhust Flame 27, 51 
(1976). 

13. Markstein, G. H.: Sixteenth Symposium 
(Internation al^ on Combustion. P. 1407, 
The Combustion Institute, 1977. 

14. Markstein, G. H. and deRis, J.: Twentieth 
Symposium (I nternational^ on 
Combustion. P. 1637, The Combustion 
Institute, 1984. 

15. Markstein, G. H.: Twentieth Symposium 
(Internation an on Combustion. P. 1055, 
The Combustion Institute, 1984. 

16. Markstein, G. H.: Radiant Emission and 
Smoke Points for Laminar Diffusion 
Flames of Fuel Mixtures, Report FMRC 
J.I. OMON4.BU, RC86-BT4, Factory 
Mutual Research Corporation, Norwood, 
Massachusetts, June 1986. 

17. Lekan, J.: Microgravity Research in 
NASA Ground-Based Facilities, NASA 
TM- 10397. NASA Lewis Research 
Center, Cleveland, Ohio, 1989. 



Section 14 

'GRAVITATIONAL INFLUENCES ON THE BEHAVIOR OF CONFINED 

DIFFUSION FLAMES" 



U. Hegde and M. Y. Bahadori 



Paper AIAA-92-0334, AIAA 30th Aerospace Sciences Meeting, 
Reno, Nevada, January 1992 




iifiiA 



AIAA 92-0334 

Gravitational Influences on the Behavior of 

Confined Diffusion Flames 

U. Hegde 

Sverdrup Technology, Inc. 

Lewis Research Center Group 

Brook Park, OH 

and 

M. Y. Bahadori 

Thermal Sciences Division 

Science Applications International Corporation 

Torrance, CA 



30th Aerospace Sciences 
Meeting & Exhibit 

January 6-9, 1992 / Reno, NV 



For permission to copy or republish, contact the American Institute of Aeronautics and Astronautics 
370 1 Enfant Promenade, S.W., Washington, D.C. 20024 



AIAA-92-0334 

GRAVITATIONAL INFLUENCES ON THE BEHAVIOR OF CONFINED DIFFUSION FLAMES 

U. Hegde* 

Sverdrup Technology, Inc. 

NASA Lewis Research Center Group 

Brook Park, Ohio 44142 

M. Yousef Bahadori" 

Science Applications International Corporation 

Thermal Sciences Division 

Torrance, California 90501 



Abstract 

Closed form, mathematical representations 
for the shapes of laminar diffusion flames stabilized 
on two-dimensional slot burners are derived for 
arbitrary gravitational levels. The solution technique 
employs the Schvab-Zeldovich conserved species 
variable. Effects of gravity on the velocity field are 
incorporated through the momentum equation; these 
effects, in turn, influence the species field. The 
results reduce smoothly to the classical Burke- 
Schumann solution as gravity approaches zero. It is 
shown that both flame height and width increase 
with decreasing gravity levels. Several interesting 
cases are considered explicitly; these include effects 
of negative and extremely high values of gravity. In 
the latter case it is shown that the flame collapses 
onto the burner surface. The analysis finds applica- 
tion in the modelling of steady and unsteady gas 
and solid propellant diffusion flames. 



a 
b 
c 
D 

g 
G 

Go 
h 

I 
M P 



Nomenclature 

one half of fuel slot width 

one half of slot burner width 

one half of burner chimney width 

binary diffusion coefficient 

gravitational acceleration 

effective acceleration, defined by Eq. (21) 

reference acceleration, defined by Eq. (22) 

flame height 

flame height at zero gravity 

defined in Eq. (18) 

mass injection rate of fuel 



* Supervisor, Microgravity Science Section. Senior 
Member AIAA. 

**■ Senior Scientist. Senior Member AIAA 
This paper is declared a work of the US Govern- 
ment and is not subject to copyright protection in 
the United States. 



M mass injection rate of (fuel + oxidizer) 

N total number of species present 

p pressure 

Pe Peclet number 

u x-component of velocity 

v y-component of velocity 

W F molecular weight of fuel 

W„ molecular weight of oxidizer 

x coordinate parallel to flame axis 

y coordinate normal to flame axis 

y, line separating hot and cold flow regions 

y, y-coordinate of flame surface 

Y Schvab-Zeldovich species variable 

Y F mass fraction of fuel 

Y. mass fraction of oxidizer 

v defined by Eq. (5) 

v P stoichiometric coefficient of fuel 

v u stoichiometric coefficient of oxidizer 

p density 

y stream function 

V, defined by Eq. (18) 

y^ summation from n = 1 to n = CO 

subscripts 

reference quantity measured at y = 

1 value in the hot flow region 

2 value in the cold flow region 



Introduction 

This paper considers the effects of gravita- 
tional fields on laminar diffusion flames stabilized 
on two dimensional slot burners in a confined co- 
flow of fuel and oxidizer. This is a classical prob- 
lem of interest and analytical solutions for flame 
shape and height in the absence of gravity are well 
known 1 - 1 . Due to the lack of available closed form 
solutions for buoyant flames, these results have 



been used in the past to describe flame shapes even 
in the presence of gravity. However, in recent years, 
it has come to be realized 3 ' 5 that the ambient gravity 
level can significantly influence observable physical 
characteristics of such flames. It is only for momen- 
tum dominated laminar flames (i.e., flames for 
which the Froude number, UoVgh, is much greater 
than unity) that the zero gravity solution has 
validity. 

Numerical and experimental studies on 
laminar gas diffusion flames where buoyancy ef- 
fects have explicitly been taken into account have 
found that gravitational fields significantly affect 
diffusion flame height and shape 1 **. Based on these 
findings it is now recognized that there are two 
opposing aspects of gravity on diffusion flame 
heights. The increase in flow velocity caused by 
buoyancy tends to increase the flame height How- 
ever, conservation of mass demands that the stream- 
lines of the flow converge. This results in an in- 
crease in species diffusion rates which contribute to 
a reduction in flame height Following this line of 
reasoning, the effect of a gravitational field on 
flame height depends upon which factor is more 
important; sometimes this may be determined by 
burner geometry. For example, from analytical 
reasoning, Roper* has derived simplified expressions 
for flame height which show that buoyantly domi- 
nated flames are shorter than zero gravity flames 
for slot burners and of the same height for circular 
burners. 

While the above noted expressions* are 
available for flame heights of burner stabilized 
buoyant flames, no closed form solutions are avail- 
able for flame shapes and widths under arbitrary 
gravity levels. Such solutions are important for a 
variety of diffusion flame applications. For exam- 
ple, radiative and conductive heat transfer from 
diffusion flames depend upon the overall flame 
shape'. In addition, the behavior of unsteady diffu- 
sion flames can also, in many cases, be understood 
utilizing simplified expressions for the steady (time 
averaged) flame shape 10,11 . 

Diffusion flame investigations also find 
application in solid propellant combustion studies. 
Solid propellant flames are known to consist of 
intricate combinations of premixed and diffusion 
flames'*. To simplify their investigations, research- 
ers have often resorted to studying simplified 
flames which retain important characteristics of 
actual solid propellant flames. For example, slot 



burner flames 11 - 11 , of the type considered in this 
paper, have been used to simulate the characteristics 
of sandwich type propellant flames 1 *. However, 
buoyant effects on such flames have largely been 
neglected. 

In this paper, analytical solutions for the 
shapes of laminar diffusion flames stabilized on slot 
burners are developed. Arbitrary gravitational fields 
directed along the axis of the flame are incorporated 
in the solution. The Schvab-Zeldovich conserved 
species formalism 2 is utilized. The conservation 
equations of mass, momentum and species are cast 
in stream function coordinates. Approximations on 
flame density and buoyant velocities, no more strin- 
gent than in previous analyses, are utilized to obtain 
closed form solutions for the conserved species 
variable. The flame shape is then obtained by the 
flame sheet approximation. These solutions reduce 
smoothly to the classical Burke-Schumann solution 
as the gravity level approaches zero. A variety of 
important cases are considered in the paper, these 
include cases with negative gravity (inverted 
flames), partial gravity and extremely high gravity 
levels. 

Formulation 

The configuration considered is shown in 
Fig. 1. Fuel flows through a slot of width 2a. 
There is a co-flow of oxidizer in the two outside 
slots. The width of an oxidizer slot is (b-a). The 
slot burner is placed in a chimney of width c 
through which there may be a secondary stabilizing 
flow. It is assumed that both fuel and oxidizer are 
injected at velocity u,,. These streams are heated by 
the presence of the established diffusion flame. The 
stabilizing (secondary) flow in the outer chimney is 
injected with an axial velocity ife and is at the 
ambient (cokf) density. Buoyancy causes the hotter 
stream above the burner to accelerate. 

The equation for the line, y,(x), separating 
the hot and cold flows is first obtained from overall 
mass conservation assuming (a) the hot flow is at a 
constant density p, and (b) the secondary flow has 
constant density p, and axial velocity u«. It will be 
shown later (see Eqns. (14) and (15)) that assump- 
tion (a) also implies that the u, is a function of x 
alone. This expression for y, is readily shown to be 



_ 2?[p 1 u 01 -p 2 u 02 ] 

1 [PiUi-PjUoj] 



(1) 



fr.^( P u,(pi»|r (7) 



An one step reaction between fuel and 
oxidizer is assumed to take place 



v P F + v Ox 



ViH 



The Schvab-Zeldovich conserved variable for chem- 
ical species is defined by 



~ ■«• F 1 ^ v i» 



y= [ ~Q* u. v 



V ox V ox &W y 



(2) 



The equation for overall mass conservation 
may be written as 



The line y,(x) from Eq. (1) transforms into 
the line 

, i= M Pl u - P2 u 02 ] 

[1-^22] 
Pl"l 



It should be noted that y, is not, in general, a 
streamline of the flow. Furthermore, assuming that 
the diffusion coefficients in the hot and cold flows 
are separately constant the species equation can be 
written for the hot and cold flows as follows 



I=(P^)(PA)S 



3 V 2 



(9) 



i£=(p 2 "o 2 ><PA>0 



(10) 



Using standard assumptions of perfect gas behavior, 
unity Lewis number and constant average molecular 
weight the conservation equation for Y is 

V dx P dy dy 9U dy (4) 

where, as in previous approaches 110 , axial diffusion 
has been neglected. 

Next, stream function coordinates are intro- 
duced by means of the two dimensional von Mises 
transformation 



•-$ 



(5) 



The above two equations are connected 
across the line y, by the conditions 



and 



(11) 



00 vl . = (Y) yl . 



(PA) 19M l% ] 

■(Pa^)(Pa^)l|r] 

(12) 

where the latter condition is obtained by integrating 
Eq.(7) across y,. 



-pv= 



=ii 



dx 



(6) 



The overall continuity equation is then satisfied and 
the species equation becomes 



Consideration is now given to the case 
where the secondary flow velocities are much 
smaller than the primary flow velocities; specifical- 
ly, cases where 



(P 2 "o2> (P 2 -P 2 ) 
<Pi"i) (PA) 



< 1 



arc considered so that it is appropriate to write 

= 



[■& 



and 



t|r x = 2^1^ (13) 



which is independent of x. In this case, y, becomes 
equal to the mass flow, M, exiting the burner and 
represents the streamline leaving the burner at y = 
b. 

The problem is now reduced to solving 
Eq.(9) subject to the following conditions 



at v = 0, M, for x > 0: 



3Y 
8y 



= 



and at x = 0: 

< v < Mp : Y = 1 

M F < V < M: Y = -(Y om , < /Y P0 )(W F v F /W„v J = -v 

The velocity u, is obtained from the axial 
momentum equation for the hot flow. Assuming 
that the pressure at any location x is imposed from 
the cold flow one obtains in stream function coordi- 
nates 

Pi u i-3J = (P 2 -Pi>^ (14) 

where the gravitational acceleration, g, is positive 
when directed along the negative x- axis. The solu- 
tion for u, becomes 



Pi 



(15) 



which is a function of x alone. 



Y(x,y) may now be solved for in the re- 
gion (x > 0, < v < yj using standard techniques. 
Upon transforming back into (x.y) coordinates the 
solution may be written as 






mi 

. n'n'i 



sin (nn—£) cos ( Kl lJ )e p «* 



where 



J = / X Pi u i<**o 



(16) 
(17) 



and Pe is the Peclet number given by 



(18) 



The solution may be extended to the region 
y, < y < c by solving Eq. (10) subject to the con- 
nection condition Eq.(Il) and the impermeable wall 
condition 



BY 



= 



at y = y, (corresponding to y = c). However, in 
general, Y(y,) will be a function of x ; hence, a 
technique such as that of Penner et al" needs to be 
employed to obtain particular analytical solutions. 
The matter is not pursued further in this paper. 



Results and Discussions 

It is readily shown that in the limit of 
vanishing g, u, = u,,, and I = p lUo ,x, so that Eq. (16) 
reduces to 

r=^- v(1 -^, +2 £_(!±v)_ 

M M ^ nit 

j,f _ n*x 2 x 

sin(nn—£)cos(mz£)e Pab 

(19) 

which agrees with the classical Burke-Schumann 
solution 1,2 . This solution is also obtained in the limit 
of p! approaching p t . 



The flame surface is obtained utilizing the 
flame sheet approximation ; that is, the condition Y 
= at the flame surface. Thus, at any axial location 
x, the coordinate y f of the flame surface is obtained 
from 



= — ?-v [1-— £] +2V 



(1+V) 



M 



M 



nn 



-n'n'I 



. M F nnp 1 u 1 y f ^^ 

sin(mt— -) cos ( * 1 f ) e P * M 

M M 

(20) 



Furthermore, the flame height, x = h, is obtained 
when y, = 0. 

It is interesting to consider the variation of 
flame height with burner width for constant burner 
flow rates. Inspection of Eq.(20) reveals that I must 
then remain constant. For momentum dominated 
flames, for x = h, I = u^h. For a constant flow rate 
buoi remains constant so that in this case h varies 
linearly with b. However, for buoyancy dominated 
flames u ; is, for practical purposes, independent of 
Uo, so that I is independent of both Uq, and b. Thus, 
for buoyancy dominated flames, the flame height is 
independent of the slot width. This recovers the 
results of Roper et al* for slot burners. 



The above equation is comparable to Eq. (5) of Ref. 
8. 

The behavior of h/h„ for non-negative 
values of G is plotted in Fig. 2. It is seen that as G 
increases, the flame height decreases. In the limit of 
very high accelerations (as G -> oO), h -> 0. 
Inspection of Eq. (16) and the initial conditions at x 
= indicates that in this case, Y can be zero only at 
the location (x,y) = (0,a) so that the flame collapses 
onto the burner. 

As G increases, the flame width, like the 
flame height, also decreases. This is shown in Fig. 
(3) where calculated flame shapes for G equal to 
Uo, 2 /b and 2u„,Vb are plotted. The stoichiometric 
coefficients chosen for this calculation correspond 
to methane burning in oxygen so thai v = 0.25. In 
addition, the physical parameters are a/b = 0.1, and 
Pe = 10. 

Care must be taken in evaluating Eqns.(16) 
and (20) for negative values of G. In these cases, 
the streamlines above the burner surface diverge 
and a stagnation plane is obtained at 



u, 



x s = 



01 



2G 



The variation of flame height with the 
gravitational level g, when the other parameters of 
the problem are kept constant may also be obtained 
from the condition that I for x = h remains 
unchanged. Define an effective acceleration 



G= _(P iI PiL 

Px 



(21) 



The developed theory is not valid in the neighbor- 
hood of the stagnation plane. A rough indication of 
the region of validity may be obtained from Eq. 
(13) which indicates that 



Yi « 



ba 



01 



u. 



and a reference acceleration 

G = UbiVAo 



(22) 



where ho is the flame height at zero gravity. Then, 
utilizing Eqns.(15), (17) and (20) it may be shown 
that 



Certainly, an upper limit to y, is the width c of the 
outer chimney so that the theory is not expected to 
be valid for 



"i < 



bu. 



01 



_^=ik[[i + 34] 2/3 - i] 



h Q 2C 



(23) 



or equivalent^ for 



X > 



■'01 



2 ((V-1] 

c 

2G 



For axial distances less than the above value of x, 
Eqns. (20) and (21) yield expressions for the flame 
shape and height Note also that Eq. (23) indicates 
that the flame height will be less than x, when 
G/G < -1/3. 

Calculated flame shapes for G equal to 
- 0.05 Uo, 2 /b and -0.42uo,Vb are shown in Fig. (4). 
The other parameters of the calculation are the 
same as those for Fig. (3). Note that the flames are 
bigger and broader than those obtained when G is 
non-negative. Calculations also show that the flame 
surface tends to spread outward from the axis for 
negative values of G having magnitudes much 
larger than 1/3. This behavior agrees with observa- 
tions of inverted flames under normal gravity condi- 
tions which exhibit tulip shapes*. 



Conclusions 

The Burke-Schumann solution for slot 
burner stabilized diffusion flames has been extended 
to account for the effects of axially directed gravita- 
tional fields. It has been shown that both flame size 
and height decrease with increase in the gravity 
level. Predictions from the developed model are 
consistent with observed behavior of momentum 
and buoyancy dominated flames. These include the 
independence of buoyancy dominated flame height 
on burner width and lateral spreading of flames 
under negative accelerations. The analysis may be 
applied to steady and unsteady gas and solid propel- 
lant diffusion flame modelling. 

Acknowledgements 

This work was supported, in part, by 
NASA Lewis Research Center under Contracts 
NAS3-25266 (UH) and NAS3-22822 (MYB). 

References 



6. Edelman, R. B. and Bahadori, M. Y., Acta Astro- 
nautica. Vol. 13. No. 11/12, pp. 681-688. 1986. 

7. Bahadori, M. Y., Edelman. R. B.. Stacker, D. P., 
and Olson, S. L.. AIAA Journal, Vol. 28, pp. 236- 
244, 1990. 

8. Roper, F. G.. Smith, C, and Cunningham, A. C, 
Combustion and Flame, Vol. 29. pp. 227-234. 1977. 

9. Bahadori, M. Y., Edelman, R. B., Sotos, R. G., 
and Stocker, D. P.. AIAA Paper No., 1991. 

10. Hegde, U. G. and Zinn, B. T., AIAA Paper No. 
89-2662, 1989. 

11. Chen, T. Y., Hegde, U. G., Daniel, B. R. and 
Zinn, B. T., AIAA Paper No. 90-3929, 1990. 

12. Beckstead, M. W., Derr, R. L., and Price, C. F., 
AIAA Journal. Vol. 8, No. 12, 1970. 

13. Kumar, R. N., Strand, L. D.. and McNamara, R. 
P., AIAA Paper No. 76-669. 1976. 

14. Price. E. W., Sambamurthi, R. K.. Sigman, R. 
K., and Panyam, R. R., Combustion and Flame, 
Vol. 63, pp. 38 1-4 13. 1986. 

15. Penner, S. S.. Bahadori, M. Y., and Kennedy, 
E. M., in Dynamics of Flames and Reactive Sys- 
tems (Eds: J. R. Bowen, N. Manson , A. K. 
Oppenheim and R. I. Soloukhin), Vol. 95, Progress 
in Astronautics and Aeronautics, pp. 261-292, 
AIAA, Washington, DC, 1985. 



1. Burke, S. P., and Schumann, T. E. W., Industrial 
and Engineering Chemistry, Vol 20, pp. 998, 1928. 

2. Williams, F. A., Combustion Theory. 2nd Ed., 
Benjamin/Cummings, CA, 1985. 

3. Edelman. R. B., Fortune, O., and Weilerstein, G.. 
Analytical Study of Gravity Effects on Laminar 
Diffusion Flames, NASA CR- 120921. 1973. 

4. Roper, F. G., Combustion and Flame, Vol. 29, 
pp. 219-226. 1977. 

5. Glassman, I., Combustion . 2nd Ed., Academic 
Press, IncFL, 1987. 



y-o 



Developed 

Diffusion 

Flame 




Figure 1. 



1.00 



Oxidizer Oxidize; 

A Schematic of the Slor Bumer Model Configuration. 




100 



Figure 2. Diffusion Flame Height, h/h„, as a Function of the Gravity Level G/G 




V/k 



Figure 3. Diffusion Flame Shapes for G = u^Vb ( — ) and G = 2u OI 2 /b ( ). 




1.0 



Figure 4. Diffusion Flame Shapes for G = -0.05 Uo, a /b ( ) and G = -0.42 u^Vb ( ). 



Section 15* 
■COMBUSTION OF GASEOUS FUELS UNDER REDUCED-GRAVITY CONDITIONS' 



M. Y. Bahadori and E. B. Edelman 



Paper LB-038, in press in Proceedings of the Second Symposium on 
Lunar Bases. Lunar and Planetary Institute, Houston, Texas, 1992 



This publication is a revised version of Paper LBS-88-014. Paper presented at the Second 
Symposium on Lunar Bases (see Section 6) 



COMBUSTION OF GASEOUS FUELS UNDER REDUCED-GRAVITY CONDITIONS 

M. Yousef Bahadori 

Science Applications International Corporation, Thermal Sciences Division 

21151 Western Avenue, Torrance, California 90501 

Raymond B. Edelman 

Rockwell International Corporation, Rocketdyne Division, WC-79 

6633 Canoga Avenue, Canoga Park, California 91303 

Abstract - The need for an improved understanding of fires is becoming critically important with increased 
space travel and utilization. While the control of fires in low-gravity environments is not well-understood, 
it is known that buoyancy significantly affects the flame behavior and characteristics. The objective of this 
research is to gain a more fundamental understanding of fires, and to quantify flame behavior under 
reduced-gravity levels. Non-premixed flames of gaseous fuels are considered in this study because they 
are relatively simple and easy to control, yet embody mechanisms found in all types of combustion 
processes ranging from uncontrolled fires to practical combustion systems. This paper presents some 
recent results from microgravity studies of these flames. In addition, the potential usefulness of Lunar- and 
Martian-based laboratories are discussed to understand the characteristics and behavior of fires in reduced- 
gravity environments. 

INTRODUCTION 

The problem of fire safety has been of equal concern both on Earth and in space. The 21st Century 
will begin a period of regular space travel, manned space stations, Lunar and Martian bases, and deep- 
space exploration. All of these activities raise the question of fire prevention in space, and the use of low- 

1 



gravity environments to further our knowledge of combustion on Earth. 

Microgravity combustion research has been vigorously pursued in the last decade in relation to fire- 
safety issues as well as the fundamental understanding of combustion phenomena. Combustion studies 
of solid, liquid, and gaseous fuels have been conducted in earthbound facilities which provide short 
durations of microgravity. Promising results have so far been obtained to warrant the continuation of this 
branch of combustion science. 

The objective of this research is to gain a more fundamental understanding of fires, and to quantify the 
flame behavior under reduced-gravity levels. Non-premixed flames of gaseous fuels are being investigated 
because they are relatively simple and easy to control, yet embody mechanisms found in all types of 
combustion processes. 

In the following sections we present: (a) general characteristics of laminar and turbulent diffusion (i.e., 
non-premixed) flames, (b) available earthbound facilities for conducting reduced-gravity combustion 
studies, (c) some new results obtained from laminar diffusion-flame studies in microgravity, and (d) critical 
need for the understanding of low-gravity turbulent flames, all directed toward the goal of understanding 
the behavior of flames not only in space but on Earth as well. 

LAMINAR AND TURBULENT DIFFUSION FLAMES 

The term "diffusion flame" classifies those types of flames in which the fuel and oxidizer are not 
premixed, whether the reactants are in the solid (e.g., coal combustion), liquid (e.g., droplet combustion), 
or gaseous form (e.g., cigarette lighter flame). Unfike "premixed" flames, as in internal-combustion engines, 
the burning process in diffusion flames is governed by diffusion of the fuel gas and oxygen toward each 
other to form a thin flame sheet which separates the two reactants. The schematic diagram of a normal- 



gravity gas-jet diffusion flame burning in a quiescent oxidizing environment is shown in Fig. 1 , where the 
different physico-chemical phenomena governing the combustion process are indicated. The gaseous fuel 
(e.g., methane) is injected through a nozzle, the tip of which acts as a flame holder. 

In combustion processes, coupling exists between chemical kinetics, fluid dynamics, diffusion of species, 
inertia, radiation, and soot formation and disposition. In addition, under non-zero-gravity condition, 
buoyancy is imposed on these processes (due to the density difference between the hot combustion 
products and the cold environment). The buoyant force causes the hot products to be removed from the 
flame in the direction opposite to the direction of the gravitational force. This complicates the understanding 
of the coupled processes involved in combustion. 

In zero-gravity environments, the buoyant force is eliminated, and the remaining processes become 
more tractable. Isolation, or even reduction of buoyancy, makes it easier to understand the interplay 
between these chemical and physical processes which are not seperable regardless of the gravity level. 
These phenomena are responsible for the very different behavior of flames observed in microgravity 
compared to those in normal gravity (Edelman and Bahadori, 1986). 

Gas-jet diffusion flames are selected in this study because they are a representative of a wide variety of 
combustion processes from the fundamental standpoint. These flames are laminar or turbulent, depending 
on the combination of jet momentum, nozzle diameter, and fuel properties. The classical behavior of a gas- 
jet diffusion flame in normal gravity (Hottel and Hawthorne, 1949) is shown in Fig. 2 which plots the 
dependence of length and structure of the flame on fuel velocity for a tube of given size. As the jet velocity 
increases, the flame transists from laminar (where mixing is governed by molecular diffusion) to fully 
developed turbulent behavior (where mixing is largely due to eddy diffusion or convection, with the final 
homogeneity being attained by molecular diffusion). It is this type of behavior which is anticipated to be 
strongly affected by the reduction in gravity level, as discussed later. 



LOW-GRAVITY EARTHBOUND AND ORBITER FACILITIES 

Figure 3 shows several earthbound and Space-Shuttle facilities which provide low-gravity environments 
for combustion research; for detailed description of these facilities, see Lekan (1989). To-date, most of the 
reduced-gravity combustion studies (including premixed flames, solid-surface combustion, laminar gas-jet 
diffusion flames, particle-cloud combustion, pool fires, and droplet combustion) have been conducted in the 
2.2-Second Drop Tower (10 5 g), 5.18-Second Zero-Gravity Facility (I0" 5 g). and Model 25 Learjet (10 2 g for 
approximately 15 seconds, attached payloads) of NASA Lewis Research Center. In addition, studies are 
planned in the KC-135 aircraft of NASA Johnson Space Center (10'^, approximately 20 seconds for 
attached payloads; 10 3 g, approximately 8 seconds for free-floating payloads). For an overview of 
combustion studies in low-gravity environments, see Sacksteder (1991 ). The middeck and spacelab of the 
Space Shuttle (1 s g) provide much longer test times and lower gravity levels, and also allow more detailed 
diagnostic measurements of flames. 

In the 2.2-Second Drop Tower, the experiment package is enclosed in a drag shield which has a low 
drag coefficient. As the drag shield falls in this 27-m tower, the experiment package is released inside the 
shield. The air drag associated with the relative motion of the package within the shield is the only external 
force acting on the package. The shield comes to rest in a sand box at the bottom of the tower. The 5.18- 
Second Zero-Gravity Facility, which provides a 132-m free-fall distance, is a 6.1-m-diameter, 145-m-deep 
steel-walled vacuum chamber at 0.01 torr. The air-tight package is decelerated in a 6.1-m-deep container 
of polystyrene pellets. Aircraft flying parabolic (Keplerian) trajectories provide longer low-gravity test times, 
but at the cost of higher gravity levels. In the Learjet. the experiment is attached to the body of the aircraft. 
The KC-135 provides the same g-level as in the Learjet for bolted-down experiments, but because of its 
size, also permits free-float packages. Intermediate acceleration levels, especially Lunar (1/6g) and Martian 
(l/3g) gravities can also be achieved in the aircraft, providing opportunities to study combustion and fluid- 
physics phenomena under these specific reduced-gravity conditions. 



MICROGRAVITY LAMINAR DIFFUSION FLAMES 

Laminar diffusion flames of hydrogen and various hydrocarbons have been studied in the 2.2-Second 
Drop Tower (Cochran and Masica, 1970; Cochran, 1972; Haggard and Cochran, 1973; Edelman et al. 
1973; Haggard, 1981; Bahadori and Stocker, 1989; Bahadori et al. 1990-a, 1990-b; for a review of earlier 
work, see Edelman and Bahadori, 1986) and the 5.18-Second Zero-Gravity Facility of NASA Lewis 
Research Center (Bahadori et al. 1990-c, 1991). 

The normal-gravity flames of these fuels, when burned in quiescent oxidizing environments, generally 
flicker (due to hydrodynamic instabilities), are yellow (due to soot emission and burn-off), and are pencil-like 
in shape (due to the presence of buoyant force). In addition, the color of these flames is not strongly 
affected by changes in either pressure or oxygen concentration. This is a consequence of strong 
entrainment of oxidizer, again, due to the effect of buoyancy (see Fig. 1). 

Figure 4 shows a normal-gravity and the corresponding microgravity flame of propane. Compared to 
laminar flames in normal gravity, those observed in microgravity are flicker-free, larger, diffuse, and rather 
globular. This is due to the absence of buoyant convection, leaving diffusion a much more dominant 
mechanism of transport. In addition, these flames are generally orange-reddish in color, which is a result 
of prolific sooting. Significant soot formation is caused by local flame suffocation, since the hot products 
of combustion accumulate in the vicinity of the flame due to the absence of buoyancy. As a result, 
continued combustion depends mainly on the diffusion of oxygen toward the flame front, causing major 
pyrolysis of the hot fuel-rich portion of the flame. As can be seen in Fig. 4, the microgravity flame appears 
to have a completely open tip. This suggests that extensive soot formation, radiative loss, cooler overall 
flame temperature, and a reduced oxygen supply contribute to extinction at the flame tip. It is quite 
possible that unburned and pyrolyzed hydrocarbons may escape through the flame tip in microgravity 
environments. 



Pressure and oxygen concentration have a significant effect on flame characteristics, color, luminosity, 
and sooting behavior in microgravity compared to normal gravity (Bahadori and Stocker, 1989; Bahadori 
et al. 1990-b). Sooting was not visible in microgravity hydrocarbon flames at 18% oxygen in nitrogen, 0.5- 
atm environments, and the flames were entirely blue, whereas their normal-gravity counterparts were 
yellow, luminous, and very similar to flames under atmospheric conditions, or even high-pressure/high- 
oxygen-concentration flames. This has a very important implication, namely, there is reduced radiative 
heating and reduced hazard of flame spread to surrounding combustible materials in low-pressure/low- 
oxygen microgravity flames. Figure 5 shows the effects of oxygen concentration on normal-gravity and 
microgravity flames. 

High-pressure and high-oxygen-concentration environments also affect the enhancement and intensity 
of burning in microgravity. Massive sooting, flame-tip opening, and extinction and soot breakthrough at the 
tip were observed even in 50%-oxygen environments. The tip-opening and soot-escape phenomena are 
unique characteristics of microgravity flames. Figure 6 shows the effects of pressure. 

Recent tests (Bahadori et el. 1991) have shown that flame radiation is an order of magnitude higher in 
microgravity compared to normal gravity for laminar gas jet diffusion flames. Enhanced soot formation, 
larger flame size, and accumulation and slow transport of the hot combustion products are the contributing 
factors. Figure 7 shows the flame radiance as a function of fuel-volume-flow rate under both normal-gravity 
and microgravity conditions. The data suggest that radiative ignition of nearby materials may be promoted 
in low-gravity environments due to the increased radiative transfer. 

A mathematical model has been developed (Edelman et al. 1973) for the study of laminar diffusion 
flames under arbitrary gravitational accelerations based on the parabolic form of the equations of motion, 
which includes the effects of inertia, viscosity, diffusion, and chemical reactions. Figure 8 shows the 
excellent agreement between the predicted and measured flame heights under both normal-gravity and 



zero-gravity conditions. We have recently applied this model to a family of methane flames under different 
gravitational levels. Figure 9 shows the non-dimensional centerline velocity vs. axial distance. Clearly, 
convective effects play a major and different role for different gravitational environments. 

The shape, color, luminosity, sooting, radiation, combustion products, and other characteristics of the 
flame can not be truly understood unless experiments are conducted (along with appropriate diagnostics) 
under the g-level of interest. Then, when combined with the theoretical analysis, these results provide a 
data base for the understanding of fires both on Earth and under reduced-gravity conditions such as those 
on the Moon and Mars, or in spacecraft environments. 

TURBULENT DIFFUSION FLAMES 

Turbulent gas-jet diffusion flames under normal-gravity conditions have been the subject of extensive 
theoretical and experimental studies for a number of decades. Figure 2 shows the classical behavior of 
a turbulent jet diffusion flame. As the jet velocity increases, the flame transists from laminar to fully 
developed turbulent behavior. For the tube size used in the flame study of Fig. 2, a velocity is reached 
where further increases in the jet velocity result in no change in flame height. This is a characteristic of 
momentum-dominated flames, i.e., flames in which buoyancy is not important. 

Much progress has been made toward the characterization of momentum-dominated turbulent flames. 
However, this is not the case for low-momentum turbulent flames characteristic of unconfined fires. In this 
case, the fire research community depends primarily on empirical results which, having been obtained 
under normal-gravity conditions, have the buoyancy effect inherently embedded within these correlations. 
When buoyancy is important (i.e., low-momentum flames, unlike Fig. 2), a constant height as a function 
of velocity is not reached in the turbulent region (see Fig. 10; Wohl et al. 1949). The mechanisms 
responsible for this behavior are far from being fully understood. Thus, the need for more fundamental data 



and analysis is apparent because of the requirement to define the hazard and control of fires not only on 
Earth but in space as well. 

For low-momentum flames, strong interactions between buoyancy and turbulent-flame structure exist 
which affect the flame behavior and chemistry through two gravity-induced mechanisms. The first arises 
directly from the buoyant force acting on the time-averaged or mean flow field, and appears as a gravity 
term in the mean momentum equation. The second mechanism arises out of the interaction between 
density and velocity fluctuations which appears as a source of turbulent kinetic energy. Under normal- 
gravity conditions, it is not possible to separate these two effects in terms of their impact on mixing rate, 
and hence, flame structure. Clearly, the advantage of operating in a low-gravity environment would be to 
provide a significant base of new information by isolating the combined effects of buoyancy. Then, with 
a theoretical model, the effects of buoyancy on the mean flow and on the generation of turbulent kinetic 
energy would be separable. 

CLOSURE 

Although the effects of buoyancy on low-momentum flames have been qualitatively observed, it is only 
recently that quantitative descriptions of the phenomena affected by gravity have been attempted. 
Understanding this phenomena is not only of fundamental interest, but it is of critical importance to fire 
safety in space as well as on Earth. Furthermore, for processing and manufacturing in space, controlled 
flames used as one form of heat source are likely to be employed. 

This paper has presented results that indicate significant effects of gravity on the flame structure. 
Moreover, it has been shown that to develop a more fundamental understanding of this phenomena along 
with a reliable prediction capability, quantitative data obtained under reduced-gravity conditions uninhibited 
by test time and size imitations are needed. The potential to obtain data from experiments conducted on 

8 



Moon and Mars offers this opportunity, one which can not be equalled in earthbound facilities . 

Acknowledgements - This work is supported, in part, by NASA Lewis Research Center, under Contract 
NAS3-22822 with Science Applications International Corporation. 



REFERENCES 

Bahadori M. Y. and Stocker D. P. (1989) Oxygen-concentration effects on microgravity laminar methane 
and propane diffusion flames. Paper presented at the 1989 Fall Technical Meeting (Eastern Section) of 
The Combustion Institute, Albany, New York. The Combustion Institute, Pittsburgh. 4 pp. 

Bahadori M. Y., Edelman R. B., Stocker D. P., and Olson S. L. (1990-a) Ignition and behavior of laminar 
gas-jet diffusion flames in microgravity. AIAA J. 28 , 236-244. 

Bahadori M. Y., Stocker D. P., and Edelman R. B. (1990-b) Effects of pressure on microgravity hydrocarbon 
diffusion flames. Paper AIAA-90-0651 , AIAA 28th Aerospace Sciences Meeting, Reno, Nevada. American 
Institute of Aeronautics and Astronautics, New York. 6 pp. 

Bahadori M. Y., Edelman R. B., Sotos R. G., and Stocker D. P. (1990-c) Temperature measurements for 
microgravity laminar diffusion flames. Paper presented at the 1990 Fall Technical Meeting (Eastern 
Section) of The Combustion Institute, Orlando, Florida. The Combustion Institute, Pittsburgh. 4 pp. 

Bahadori M. Y., Edelman R. B., Sotos R. G., and Stocker D. P. (1991) Radiation from gas-jet diffusion 
flames in microgravity environments. Paper AIAA-91-0719, AIAA 29th Aerospace Sciences Meeting, Reno, 



Nevada. American Institute of Aeronautics and Astronautics, New York. 4 pp. 

Cochran T. H. (1972) Experimental Investigation of Laminar Gas Jet Diffusion Flames in Zero Gravity . 
NASA TN D-6523. NASA Lewis Research Center, Cleveland. 26 pp. 

Cochran T. H. and Masica W. J. (1970) Effects of Gravity on Laminar Gas Jet Diffusion Flames . NASA 
TN D-5872. NASA Lewis Research Center, Cleveland. 25 pp. 

Edelman R. B. and Bahadori M. Y. (1986) Effects of buoyancy on gas-jet diffusion flames: Experiment and 
theory. Acta Astronautica . 13, No. 11/12, 681-688. 

Edelman R. B., Fortune O. F., Weilerstein G., Cochran T. H., and Haggard J. B. (1973) An analytical and 
experimental investigation of gravity effects upon laminar gas jet diffusion flames. Fourteenth Symposium 
(International) on Combustion , pp. 399-412. The Combustion Institute, Pittsburgh. 

Haggard J. B. (1981) Forced and Natural Convection in Laminar Jet Diffusion Flames . NASA TP-1841. 
NASA Lewis Research Center, Cleveland. 21 pp. 

Haggard J. B. and Cochran T. H. (1973) Hydrogen and Hydrocarbon Diffusion Flames in a Weightless 
Environment . NASA TN D-7165. NASA Lewis Research Center, Cleveland. 28 pp. 

Hottel H. C. and Hawthorne W. R. (1949) Diffusion in laminar flame jets. Third Symposium on Combustion , 
pp. 254-266. Williams and WilkJns Co., Baltimore. 

Lekan J. (1989) Microgravity research in NASA Ground-Based Facilities . NASATM-101397. NASA Lewis 
Research Center, Cleveland. 16 pp. 

10 




DIFFUSION 
RAOtATlON 



FLAME STABIIZATION 



Figure 1 . Physical and chemical processes occuring in laminar gas-jet diflusion flames of hydrocarbons 
in normal-gravity environments. 




Fully developed 
turbulent flames 



KJTEnvelopei 
k flame \ 




lr-V-^-¥— - 




Increasing noule velocity — » 
Figure 2. Change in the flame height and behavior with increase in nozzle velocity for a typical gas-jet 
diffusion flame; reproduced from Hottel and Hawthorne (1949). 



Sacksteder K. R. (1991) The implications of experimentally controlled gravitational accelerations for 
combustion science. Twenty-Third Symposium (Intern ational) on Comhiistinn in press. The Combustion 
Institute, Pittsburgh. See also: NASA Lewis Research Center (1989) Microoravitv Combustion Sr. n. Q 
A Program Overview. NASA TM-1 01424. NASA Lewis Research Center, Cleveland. 18 pp. 

Wohl K., Gazley C, and Kapp N. (1949) Diffusion flames. Third Symposium on Combustion pp. 288-300. 
Williams ans Wilkins Co., Baltimore. 



11 







(b) 





//tetli 


1 


/ 


F^ff 






fc*JB 






B3c3^ 


1 


=. 


-CTS 


P. ^* 






y^y 


MMJ«. 




i_ h 




h-r 



Mil J 





Figure 3. Earthbound and Shuttle orbiter facilities which provide low-gravity environments for combustion 
research; (a) 2.2-Second Drop Tower, (b) 5.18-Second Zero-Gravity Facility, (c) Learjet. (d) KC-135 
aircraft, (e) Shuttle middeck, and (f) Spacelab. 



scile: 



1 cm 





Figure 4. Normal-gravity and microgravily flames of propane burning in quiescent air at 1 atm; nozzle 
radius = 0.0825 cm and fuel-flow rate - 1.0 cc/sec. The various colors observed are as follows: B (blue), 
BB (bright blue). D (dark). DB (dark blue). DP (dark pink). DR (dull red), O (orange). P (pink). R (red). W 
(white), and Y (yellow). The range of flicker (f) is also shown for the normal-gravity flames; reproduced 
from Bahadori et al. (1990-b). 



Propane Flame 
18S Oxygen 



Propane Flame 
211 Oxygen 




Figure 5. Effects of oxygen concentration on normal-gravity and microgravity flames of propane at 1 atm; 
nozzle radius - 0.074 cm and fuel-flow rate « 0.96 cc/sec. The various colors indicated in the diagram are 
as follows: B (blue). D (dark). DR (dull red). O (orange), P (pink). R (red), W (white), and Y (yellow). The 
bars show the range of normal-gravity flame ficker (f); reproduced from Bahadori and Stocker (1989). 



1-9 
1.5 «tm 




Vf/M 




scale: 



1 en 




Ulf 




tun 



Figure 6. Effects of pressure on normal-gravity and microgravity flames of propane burning in quiescent 
air (21% oxygen in nitrogen). For details, see Fig. 4; reproduced from Bahadori et al. (1990-b). 




0.5 1.0 1.5 
Volume Flow Rite (c» 3 /sec) 

Figure 7. Radiance as a function of fuel volume-flow rate for propane flames burning in air at 1 .0 atm; 
nozzle radius = 0.0825 cm; reproduced from Bahadori et al. (1991). 



lOOO r- 



100 



^° 



10 



Theoretical 
prediction 

Zero -g — , 
Normal-; — 1\ 




Experimental data 



Normal-? 


/■„. cm 


O 


0.051 


A 


0.0825 


a 


0.113 





0.166 


o 


0.318 


V 


0.442 



Solid symbols denote 
wo-? 



1000 



Figure 8. Comparisons between the theoretical predictions (Edelman et el. 1 973) and experimental results 
(Cochran and Masica, 1970; Cochran, 1972) for non-dimensional flame height (height / nozzle radius) vs. 
jet Reynolds number (jet velocity x nozzle radius / fuel kinematic viscosity); methane-air flames at 1 .0 atm; 
reproduced from Edelman et al. (1973). 



s 


a / 


i i i i i — r— i — r- » « 


4 




2/5 9^^ 


} 




1/* 1 ^_^^- 


o 

3 
3 




' 10-1^^ 


1 [// 


. 




io- *l 






io- J « yio-«i /oi 



4 I 12 M 20 24 » 

I (c) ► 

Figure 9. Predicted non-dimensional centerline velocity (with respect to jet exit velocity) vs. axial distance 
along the jet as a function ot gravitational level; methane-air flames, nozzle radius - 0.0825 cm, fuel-flow 
rate - 1.0 cc/sec, pressure = 1.0 atm, and jet exit velocity - 46.8 cm/sec; reproduced from Edelman and 
Bahadori (1986). 




200 400 600 800 1000 1200 1400 1600 1800 2000 

Volim flow rate, cirVs 

Figure 10. Effects of fuel-volume flow rate and tube diameter on flame height for city gas diffusion flames; 
reproduced from Wohl et al. (1949). 



Section 16 

•EFFECTS OF OXYGEN CONCENTRATION ON RADIATION FROM 
MICROGRAVITY LAMINAR PROPANE DIFFUSION FLAMES" 



M. Y. Bahadorl, R. B. Edelman, R. G. Sotos, D. P. Stocker, and D.F. Vaughan 



1992 



EFFECTS OF OXYGEN CONCENTRATION 
ON RADIATION FROM MICROGRAVTTY 
LAMINAR PROPANE DIFFUSION FLAMES 

M. Y. Bahadorif and R. B. Edelmantt 

Science Applications International Corporation 

Thermal Sciences Division 

Torrance, California 90501 USA 

and 

R. G. Sotos, D. P. Stocker, and D. F. Vaughan§ 

NASA Lewis Research Center 

Space Experiments Division 

Cleveland, Ohio 44135 USA 

Abstract - Laminar diffusion flames of propane have been studied 
under both normal-gravity and (5.18-second) microgravity conditions. 
The quiescent oxidizing environment was at 1.0 atm pressure, and had 
15% - 30% by volume of oxygen in nitrogen. Flame radiation was 
measured using a wide-view-angle, thermopile-detector radiometer. 
The results show that radiation levels are significantly higher in 
microgravity than in normal gravity. This is mainly due to the larger 



t Author and address for correspondence: M. Yousef Bahadori, SAIC/NASA LeRC, 
MS 500-217. Cleveland, Ohio 44135. 

tt Present address: Rockwell International Corporation, Rocketdyne Division, Canoga Park, 
California. 

* Baldwin-Wallace College, Berea, Ohio. 



flame size and accumulation and slow transport of the combustion 
products in the vicinity of the flame, which are caused by the removal 
of buoyancy. In addition, it is shown that the sooting characteristics are 
different in microgravity compared to normal gravity. For the normal- 
gravity flames, the effects of oxygen concentration on radiation is 
shown to be negligible, whereas in microgravity environments, flame 
shape, color, luminosity, and radiation are strongly affected by the 
amount of oxygen in the environment. This has direct application to the 
problem of spacecraft fire detection and fire safety. The data suggest 
that longer test times are needed to reach steady state in microgravity 
environments. 



1. Introduction 

The problem of fire safety in spacecraft environments has been a 
major concern. In order to understand the behavior and characteristics 
of flames in low gravity, different programs have been underway to 
study the combustion processes of solid, liquid, and gaseous fuels in 
microgravity environments. 1-4 The focus of this research is on the 
fundamental understanding of the structure of laminar gas-jet diffusion 
flames under microgravity conditions. The basic mechanisms which 
control the diffusion flames include the coupled processes of mixing, 
chemical kinetics including soot formation and disposition, flame 
radiation, and buoyancy-induced convection. A major problem 
impeding our understanding of flame behavior is the masking effect of 
buoyancy on flames in normal-gravity environments. The buoyancy- 
induced convection masks the processes of chemical-diffusional 
interactions that are fundamental to the understanding of combustion 
phenomena. In microgravity environments, the buoyant force is nearly 
eliminated, and the remaining physico-chemical processes are 
responsible for the very different behavior of flames observed in the 
microgravity experiments conducted to-date. 5 

In recent studies of microgravity laminar diffusion flames, 6 * 12 some 
unique and, in most cases, unexpected phenomena have been observed. 
The results show that flame characteristics, color, luminosity, and 
sooting behavior are strongly affected in microgravity compared to 
normal gravity. In addition, it was shown that radiation can be up to an 



order of magnitude higher in microgravity flames. Also, pressure and 
oxygen concentration have a more significant effect on microgravity 
flames than those in normal gravity. 7 * 8 

Radiation from laminar and turbulent gas-jet diffusion flames in 
normal-gravity environment has been measured by Markstein in an 
impressive and systematic way. 13 " 18 In this paper, we have followed 
his approach to measure the total radiation from microgravity flames of 
propane burning under different oxygen-containing environments, and 
have compared the results with their normal-gravity counterparts. 



2. Experimental 

The experiments were conducted in the 5.18-Second Zero-Gravity 
Facility of NASA Lewis Research Center. 19 Figure 1 shows a schematic 
of the hardware. The volume of the combustion chamber was 0.087 m 3 . 
The fuel nozzle (protruding 15 cm inside the chamber) had a tapered 
tip with inner radius of 0.0825 cm, and was positioned along the 
centerline of the cylindrical chamber. The nozzle was approximately 
20 cm away from the chamber wall. The tip of the nozzle was 
approximately 50 cm away from the chamber top. The fuel was 
supplied from a stainless steel bottle of 75 cm 3 in volume at an initial 
pressure of 100 psig. Appropriate windows were mounted in the 
chamber wall for the movie cameras and the radiometer. A pressure 
transducer was used to record the chamber pressure during the test. 

Both normal-gravity and microgravity flames of propane 
(99.9%-pure) were studied in quiescent oxidizing environments of 15% 
to 30% by volume of oxygen in nitrogen. The chamber pressure was 
1.0 atm, and the fuel-volume flow rate was 1.5 cm 3 /sec, corresponding 
to a cold-jet Reynolds number of 90. 

Most of the flames discussed here were ignited using a retracting, 
nichrome-wire, ignitor which was located 1.0 cm above the nozzle tip. 
It was removed from the flame area in 0.5 sec following ignition. Two of 
the flames discussed here were ignited using a spark ignition device 
which has since been replaced by the nichrome wire because of its 



reliability. The ignition system was activated for 0.5-1.0 sec for the 
normal gravity and microgravity tests. The microgravity flames were 
ignited after the release of the package in the drop tube. Two movie 
cameras (filming rates = 12 and 30 frames/sec, respectively) using color 
Kodak Ektachrome Video News film recorded the flame development 
and behavior. 

The wide-view angle radiometer was comprised of a 1.5 x 1.5-mm, 
fast-response thermopile detector with a Calcium Fluoride (CaF 2 ) 
window (Oriel Corporation, Model 7106), and a time constant of 
40 msec, field of view of 90 - 100 degrees, and maximum irradiance of 
0.2 W/cm 2 . In addition, the experiment chamber was equipped with a 
CaF 2 window to allow maximum transmission of the infrared radiation. 
The window had a usable transmission range of 0.2 - 9.0 ^m, with its 
absorptivity falling off above 1.0 u.m to about 50% at 10.0 urn; its 
approximate transmission in the linear range was 94%. The radiometer 
was calibrated using a black-body source (Barnes Engineering, Model 
11-210), with an opening of 2.54 cm in diameter. A nanometer 
(Keithly Instrument, Model 181) was used to measure the radiometer 
response to various temperatures of the black-body source. The 
distance from the nozzle centerline (and for calibration, from the 
black-body source) to the radiometer sensor was 28.6 cm, and from the 
shutter iris to the sensor was 3.84 cm. The iris on the shutter had an 
open diameter of 2.54 cm. The inner diameter of the Calcium Fluoride 
window was 4.45 cm. The nozzle/window/iris/sensor arrangement was 



such that the field of view on a plane passing through the centerline of 
the nozzle had a radius of 9.55 cm. 



3. Results 

Figure 2 shows the normal-gravity and microgravity flames injecting 
propane into oxidizing environments of 15%, 18%, 21%, and 30% by 
volume of oxygen in nitrogen. Following ignition, the normal-gravity 
flames establish quickly (in less than 0.5 sec), and show the typical 
flicker characteristics (with a frequency of 10 - 20 Hz) of laminar 
flames burning in quiescent environments. Both minimum and 
maximum flame heights are shown in Figure 2 for the normal-gravity 
flames. However, the microgravity flame heights continuously increase 
after ignition, and the color of these flames constantly change, indicating 
that the flame needs longer time to reach a steady state. The 
microgravity flames shown in Figure 2 are for the last instant of the 
drop (just prior to the beginning of the deceleration period). 

Figure 3 shows the flame radiance as a function of time for both 
normal-gravity and microgravity flames. Using the calibration chart, 
the data obtained from the radiometer reading (in mV) was converted 
to the equivalent black-body temperature (T b ) of the calibration source. 
The flame radiance (oT 4 b /7l) in W/cm 2 -sr was then obtained as a 
function of time, where o is the Stefan-Boltzmann constant. 

Table 1 shows the data obtained in terms of flame height, flame 
maximum diameter, radiance, and radiance per unit projected flame 
area. This area was calculated by multiplying the flame height by the 
maximum flame diameter, to give a rough estimate of the relative 



values of radiance per unit flame area between the normal-gravity and 
microgravity flames. The radiance data are for the last instant (at 
approximately 5 seconds) of the test. The heights for normal-gravity 
flames in Table 1 are the averages of minimum and maximum 
flickering-flame height. The heights for microgravity flames were 
measured as the points where blue was no longer visible on the film 
(for the 15%-0 2 -flame), and where red could no longer be seen (for the 
other flames); see Figure 2. Table 1 also shows the rate of change of 
radiance with time during the last second of the experiment for both 
normal-gravity and microgravity flames. 

In Figure 4, flame radiance as a function of oxygen concentration is 
presented. The normal-gravity flames show no significant changes in 
radiance with increasing oxygen percent. However, the microgravity 
flames show a linear variation of radiance with oxygen concentration. 



4. Discussion 

The flames of Figure 2 show the significant effects of buoyancy on 
shape, luminosity, and sooting characteristics. The normal-gravity 
flames are yellow, flickering, and pencil-like, regardless of oxygen level 
in the environment. However, their intensity appears to increase at 
higher oxygen concentrations. The microgravity flames are in general 
taller, wider, and sootier than the normal-gravity flames. The flame of 
15% oxygen is entirely blue, indicating that sooting is reduced, and 
glowing and burning particles are not present. However, the continuous 
release and slow transport of the hot combustion products due to the 
absence of buoyancy may contribute to slow heating of the fuel rich 
region of the jet, and given enough time in microgravity, the blue flame 
may start to show an increase in sooting and subsequent change in the 
color of the flame. The flame in 18% oxygen shows a transition from 
non-sooting to the highly sooting condition. As the oxygen level 
increases, the tendency for sooting increases, which is accompanied by a 
brighter flame due to higher temperatures and increased sooting. 

The orange/reddish color of the microgravity flames is an indication 
of cooler temperatures than is expected for typical burn-off conditions. 
The yellow/orange/red/dull red regions in the flames show that the 
temperature of the soot decreases in microgravity flames, leading to 
local extinction at the flame tip. This tip-opening and underventilated- 
type behavior is a characteristic of sooty microgravity flames. Absence 
of buoyancy and dominance of diffusive processes result in longer 



10 



residence times, enhanced soot formation, increased radiative loss and 
additional cooling by pyrolysis, which are, in some combination, 
responsible for the observed characteristics of microgravity flames. 

The data presented in Figure 3 show that radiative loss in normal- 
gravity flames is not only constant throughout the duration of the test, 
but that the level of radiation does not change with increasing oxygen 
concentration. This is largely due to the instantaneous removal of the 
hot combustion products under the influence of buoyancy. In addition, 
the normal-gravity flames become established immediately following 
ignition, whereas in microgravity, a continuous development of the 
flame is observed (both visually by examining the films and through 
radiation data). The data of Figure 3 also show that at lower oxygen 
levels, radiation approaches a possible steady-state value more slowly 
compared to radiation from high-oxygen-concentration flames. This is 
due to the reduced temperature and concentration gradients and, hence, 
transport rates, at the reduced ambient oxygen levels. The rates of 
change of the radiation loss are shown in Table 1 where it is apparent 
that the microgravity flames have not reached a steady-state after 5 
seconds. Table 1 also shows that for microgravity flames the radiation 
decreases with reduced ambient oxygen concentrations. The radiation 
from the normal gravity flames is affected by flickering and, 
consequently, shows no specific trend for the range of conditions tested. 
However, it is interesting to note that the radiation from these normal 
gravity flames falls within a relatively narrow band, as can be seen in 
Figure 4. Finally, Table 1 shows a non-monotonic flame length 



li 



dependence on oxygen concentration for both the normal gravity and 
microgravity flames. This may be due to the onset of chemical kinetic 
quenching caused by the lower oxygen concentrations and the attendant 
reduced temperatures. Concentration measurements, to be made in the 
future, will help to confirm the hypothesis that these flames involve 
incomplete combustion. 



12 



<y Conclusion 

The effects of ambient oxygen concentration on radiation from 
normal-gravity and microgravity laminar propane diffusion flames 
have been measured using a wide-view-angle radiometer. The results 
show that while normal-gravity flames do not exhibit appreciable 
effects of oxygen concentration on radiation, microgravity flames show 
a continuous increase in radiation with increasing oxygen concentration. 
Although these flames had not reached steady-state, the level of 
radiation from the microgravity flames was up to an order of magnitude 
larger than that from normal-gravity flames. Further quantitative data 
and analysis is underway to more fully characterize and understand the 
behavior of gas-jet diffusion flames in microgravity environments. 



Acknowledgments 

This work was supported by NASA Lewis Research Center under 
Contract NAS3-22822. 



13 



References 

1. Sacksteder, K. R.: Twenty-Third Symposium (International) on 
Combustion, P. 1589, The Combustion Institute, 1990. 

2. Law, C. K.: Combustion in Microgravity: Opportunities, Challenges, 
and Progress, Paper AIAA-90-0120, AIAA 28th Aerospace 
Sciences Meeting, Reno, Nevada, January 1990. 

3. Faeth, G. M.: Homogeneous Premixed and Non-premixed Flames in 
Microgravity: A Review, IKI/AIAA Microgravity Science 
Symposium, Moscow, May 1991. 

4. Williams, F. A.: Microgravity Sci. Technol. Ill 3, 154 (1990). 

5. Edelman, R. B. and Bahadori, M. Y.: Acta Astronautica 13, 681 
(1986). 

6. Bahadori, M. Y. and Stocker, D. P.: Oxygen-Concentration Effects on 
Microgravity Laminar Methane and Propane Diffusion Flames, 
Eastern States Meeting of The Combustion Institute, Albany, New 
York, October/November 1989. 



! 5 PRECEDING PAGE BLANK NOT FILMED 



7. Bahadori, M. Y M Stocker, D. P., and Edelman, R. B.: Effects of 
Pressure on Microgravity Hydrocarbon Diffusion Flames, Paper 
AIAA-90-0651, AIAA 28th Aerospace Sciences Meeting, Reno, 
Nevada, January 1990. 

8. Bahadori, M. Y., Edelman, R. B., Stocker, D. P., and Olson, S. L.: 
AIAA J. 28, 236 (1990). 

9. Stocker, D. P.: Size and Shape of Laminar Burke-Schumann 
Diffusion Flames in Microgravity, Central States Meeting of The 
Combustion Institute, Cincinnati, Ohio, May 1990. 

10. Bahadori, M. Y.: An Analytical Solution for Transient, Cylindrically 
Symmetric Laminar Diffusion Flames in the Absence of Buoyancy, 
Central States Meeting of The Combustion Institute, Cincinnati, 
Ohio, May 1990. 

1 1 . Bahadori, M. Y., Edelman, R. B., Sotos, R. G., and Stocker, D. P.: 
Measurement of Temperature in Microgravity Laminar Diffusion 
Flames, Eastern States Meeting of The Combustion Institute, 
Orlando, Florida, December 1990. 

12. Bahadori, M. Y., Edelman, R. B., Sotos, R. G., and Stocker, D. P.: 
Radiation from Gas-Jet Diffusion Flames in Microgravity 
Environments, Paper AIAA-91-0719, AIAA 29th Aerospace 
Sciences Meeting, Reno, Nevada, January 1991. 



16 



13. Markstein, G. H.: Fifteenth Symposium (International) on 
Combustion, P. 1285, The Combustion Institute, 1974. 

14. Markstein, G. H.: Combust. Flame 27, 51 (1976). 

15. Markstein, G. H.: Sixteenth Symposium (International) on 
Combustion, P. 1407, The Combustion Institute, 1977. 

16. Markstein, G. H. and deRis, J.: Twentieth Symposium 
(International) on Combustion, P. 1637, The Combustion Institute, 
1984. 

17. Markstein, G. H.: Twentieth Symposium (International) on 
Combustion, P. 1055, The Combustion Institute, 1984. 

18. Markstein, G. H.: Radiant Emission and Smoke Points for Laminar 
Diffusion Flames of Fuel Mixtures, Report FMRC J.I. OMON4.BU, 
RC86-BT-4, Factory Mutual Research Corporation, Norwood, 
Massachusetts, June 1986. 

19. Lekan, J.: Microgravity Research in NASA Ground-Based Facilities, 
NASA TM- 10397, NASA Lewis Research Center, Cleveland, Ohio, 
1989. 



17 



m.n.mum Lgmaxim.m rn.crograv.ly 



1 5% 2 



1 8% O 



21% O 



30% O. 




Fig. 2 



Normal-gravity (both minimum and maximum heights of 
flicker) and microgravity flames of propane burning in 
different oxygen environments. Reproduction has slightly 
changed the intensity of the flames, and has reduced the 
visible length of the blue flame. These photographs are 
included to provide qualitative information on flame color 
and luminosity only. 



RADIOMETER 




MOVIE CAMERA 



SPAM ELECTRODE 



FUEL NOZZLE 



Fig. 1 Experiment package for the 5.18-second microgravity tests. 





(O IT) T rO CN 

(js z uio/m) aouoipoy 



CO lO T ^ *N »- O 

(js z uj3/m) souoipoy 



s 



> 

I- 
00 

o 
o 

i 

C 

a 



> 
at 

i- 

i 

e 

u. 
O 

c 





r 




i 


1 — ' 


1 — I 


/ 


- 






i 

o 






o». C: 
i / 

1—1 : ?: 


- 




o 


u 






: I - 


- 




• 1 

«— < 








Ji ; 






* 


\ 


i 


' i— — l ' — ( 



(J 



£ 



io *n ■* •*> «n 

(jsjuuo/m) aouojpoa 




O 






CO 

> 

O 

c 

"3 

as 



©b 



<o tn •»'*■> cm 

(js 2 oio/m) aouojpoy 






o 

c 



7.0 



6.0 



5.0 _ 



4.0 - 



3.0 - 



2.0 - 



IT} 

<* 1.0 





Propane 










■" 


1 atm 










- 








o-g ^^^> 




- 








l-g 




-+- 


i i i 


' > 


i i 


i i i i i t 


— : — 1_ i i 



15% 



18% 21% 



30?; 



Oxygen Concentration 



Fig. 4 Flame radiance (at approximately the last instant of the test) 

as a function of oxygen concentration. 









E 






C0 












u, 






t> 


o 




c 


^.4 




« 






a, 


*■» 




o 


rt 




•- 






a. 






.2 


C 

E 

c 






o 




Q 


> 
c 




>» 


w 




+-* 






> 


00 

c 




w. 


N 




01) 






O 


•o 




t« 


• ^ 




o 


X 




s 


O 




T3 


a 




C 


o 




(4 


00 






3 


D 


OS 




w. 


h» 


s 


D 


a 


"■" 


JS 


• 


00 


a 


«9 


c 


C/5 


E 


e 


o 
E 


o 


3 


** 


£ 


a 


< 



3 



V) 

a> 

E 
ca 

u_ 

>. 

"> 
co 

o> 

o 

k. 

o 

is 


dR/dt 
at 

sec. 


m 
o 

d 


d 


CO 

d 


CO 

d 


'c 

■ 3 ® m 

"o w. E 5 
« o c5 £ 

DC Q.3= CO 


in 
d 


in 
d 


CM 

d 


CO 

d 


1 

tN 

2 E 
E . .P 

co "U ;=*> —* 


CO 
CO 

CO 


CO 


C\i 

4 


CO 

•la- 
in 


. CD 

Sill 


CO 


C\j 


c\i 


in 


CD £ 


m 


•** 
cri 


m 
oS 


CO 


CO 
CD 

E 
ca 

u. 

1 

O 

I 

"3 

E 

k_ 

o 

Z 


Cl *- Y, CD 
T3 CO JL co 


o 


o 


o 


o 


■o ! E « 

« CD I 2 
OC Q.5= CO 


C\J 

d 


00 

o 
d 


CO 

d 


in 

CM 

d 


jb tr E 
E ^r^> 
« "2 > -^ 

C 2> lo 


CO 
CO 

d 


in 

CO 

d 


3 

d 


in 

CO 

d 


. CD 

slit 


CO 

d 


d 


CD 

d 


CO 

d 


o» E .2> c 


CD 
CO 


CD 
CO 


CO 

in* 


tT 

4 


o" 

0** 


in 


CD 


cn 


3* 
o 

CO 



Section 17 

•PRELIMINARY OBSERVATIONS ON THE EFFECTS OF BUOYANCY ON 
TRANSITIONAL AND TURBULENT DIFFUSION FLAMES" 



M. Y. Bahadori, D. F. Vaughan, D. P. Stocker, K. J. Weiland, and R. B. Edelman 



Paper presented at the Central States Meeting of The Combustion Institute, 

Columbus, Ohio, April 1992 



PRELIMINARY OBSERVATIONS ON THE 

EFFECTS OF BUOYANCY ON TRANSITIONAL AND 

TURBULENT DIFFUSION FLAMES* 

M. Yousef Bahadorif 

Science Applications International Corporation 

Torrance, California 

David F. Vaughantt- Dennis P. Stocker, and Karen J. Weiland 

NASA Lewis Research Center 

Cleveland, Ohio 

and 

Raymond B. Edelman§ 
Rockwell International Corporation, Rocketdyne Division 

Canoga Park, California 



Abstract 

The investigation of gas jet diffusion flames of propane under both normal-gravity and 
microgravity conditions has been extended from the laminar region to the transition-to-turbulent and 
the fully developed turbulent regimes. Due to the significant reduction in the influence of buoyancy- 
generated turbulence in microgravity, flames burning in air at atmospheric pressure show different 
characteristics in microgravity compared to their normal-gravity counterparts. These characteristics 
include flame height, color, stand-off distance, luminosity, and sooting. Flame behavior as a. function 
of jet Reynolds number for the three regimes of laminar, transitional, and turbulent is discussed. The 
results show that the microgravity flame height increases smoothly throughout the laminar region to the 
turbulent regime, and is longer at all flow rates than the normal -gravity flame. 



* 



Presented at the 1992 Spring Technical Meeting of The Combustion Institute, Columbus, Ohio, 
April 26-28, 1992. 



t Author and address for correspondence: M.Y. Bahadori, SAIC/NASA LeRC, MS 500-217, 
Cleveland, Ohio 44135. 

ft Student presenter, Baldwin-Wallace College, Berea, Ohio. 

§ Consultant 



1. Introduction 

A fundamental understanding of flame structure and characteristics under reduced-gravity 
conditions is essential to the problem of fire safety in spacecraft environments.* 1 ) A major problem 
impeding our understanding of flame behavior in normal gravity is the masking effects of buoyancy, 
which results in significant differences between the structure of normal-gravity and microgravity 
flames.* 2 ) 

In recent studies of microgravity laminar diffusion flames, some unique and unexpected 
phenomena have been observed.* 23 ) The results have shown that flame color, size, luminosity, 
sooting behavior, approach toward steady state and radiative loss, as well as the effects of pressure 
and oxygen concentration on these characteristics, are significantly different in microgravity compared 
to normal gravity. 

No previous study exists on the influences of the removal of buoyancy on transitional and 
turbulent gas jet diffusion flames. The present work characterizes the different regimes of flame 
behavior under both normal-gravity and microgravity conditions. The available microgravity time and 
the experiment volume restrict the study to fairly small diffusion flames in the laminar, transitional and 
turbulent regimes. 



2. Experimental 

Diffusion flames of propane burning in quiescent air at atmospheric pressure were studied in the 
2.2-Second Drop Tower of NASA Lewis Research Center. The combustion chamber was an open, 
rectangular (26 x 29 x 61 cm 3 ) container. The chamber and accessories where placed inside the 
(closed) drag shield. One side of the chamber was made of Lexan to provide photographic access. 
The other three (steel) sides were painted black to prevent the reflection of light from the walls. The 
vessel was covered by a finely woven copper screen to prevent flame-tip disturbance. 

The fuel nozzle had a tip inner radius of 0.039 cm. A 4.4-cm elevated aluminum plate held the 
nozzle above the chamber floor. The plate, measuring 19 cm by 20 cm, acted as a flame holder by 
reducing the entrainment of the surrounding air. The nozzle, located at the center of the plate, 
protruded 2-3 mm above the plate to prevent water-vapor condensation. 

Ignition was achieved using a hot-wire system. The ignitor was activated for up to 1 second. 
Once ignited, the hot wire was retracted from the flame. Ignition was achieved in normal gravity. Up 
to 4 seconds were allowed for the flame to establish itself and to film the normal-gravity flame prior to 
the release of the experiment into free fall. The fuel flow was maintained for the duration of the 
2.2-second microgravity test 

A 1 6-mm Milliken movie camera recorded the normal-gravity and microgravity flames at a filming 
rate of 50 frames/second. Tungsten Eastman Ektachrome Video News films were used to record the 
flames. 

The fuel was 99.9%-pure propane. The volume-flow rates were 100 to 1000 cm 3 Anin, 
corresponding to a cold-jet Reynolds number of 314-3140, based on the nozzle radius. 



3. Results and Discussion 

Figure 1 shows the variation of flame height with (cold jet) Reynolds number. A study of this 
diagram reveals some interesting and unexpected characteristics. 

The normal -gravity flames of the laminar regime show a linear variation of height with Reynolds 
number, as is expected from the classical behavior.( 6 ) These flames flicker due to the hydrodynamic 
instability, and the flicker range increases with increasing jet momentum. The laminar flames are 
generally yellow with a small blue base. As the jet momentum increases, the (average) flame height 
undergoes a transitional regime with a dip which is not a characteristic of typical transitional flame/ 6 ) 
However, other studies (e.g., 7) have shown that relatively low-momentum flames exhibit this kind 
of behavior and, indeed, the flame height in the turbulent regime can be larger than that in the 
transitional regime. 

In the transitional regime of normal-gravity flames, the flame starts to lift off, having a major blue- 
base section followed by a yellow brush which both flickers and wavers from side to side. This 
behavior persists throughout the turbulent regime, although the boundary of the break point shows the 
classical behavior. Blow-off occurs somewhere beyond Re = 3,000. 

The micTOgravity flames, on the other hand, show a completely different behavior. In the laminar 
regime, the flame is flicker-free, much wider than the normal-gravity flame (up to four times larger in 
radius), and has an open tip with flame colors changing from blue at the base to yellow, then orange, 
then red, and finally dark red toward the flame tip. The flame color and shape indicate soot quenching 
and its subsequent escape through the tip, which resembles an underventilated-type behavior. The 
flame-tip opening in addition to the criteria for measuring the height of open-tip flames have been 
discussed in detail elsewhere.* 4 ) However, as the jet momentum increases toward the transitional 
regime, the width of the flame tip decreases. In addition, the gap between microgravity and (average) 
normal-gravity flame heights increases with increasing Reynolds number. 

The microgravity flames (at least for those particular Reynolds numbers of this study) show a 
monotonic flame-height variation, which means that the transition from the laminar region to the 
turbulent regime may occur quite smoothly, unlike the classical behavior for the normal-gravity flames. 
However, somewhere in the transition regime, the flame tip closes and large-scale, slow-moving 
structures develop which do not look like a brush-type behavior, as in the normal-gravity flame, but 
rather a wrinkled structure. The flame lifts off the nozzle tip, although the stand-off distance is almost 
half of that for the normal-gravity flame. Tip flicker was not observed for any of the microgravity 
flames. The flame maintains a constant length up to a Reynolds number of 2200, and then increases in 
length, which unfortunately falls out of the camera field of view. It is anticipated that the height will 
increase in this region with increasing Reynolds number. The microgravity flame height in the 
turbulent regime is almost twice that of the corresponding normal-gravity flame. 



4. Conclusions 

Diffusion flames of propane burning in air at atmospheric pressure have been studied under both 
normal-gravity and microgravity conditions. Laminar, transitional, and turbulent regimes have been 
investigated. The results show significant differences in the structure and characteristics of 
microgravity flames in the three regimes, compared to their normal-gravity counterparts. Work is 
underway to characterize the details of these regimes and to expand the da tab ase. 



Acknowledgements 

This work was supported by NASA Lewis Research Center under Contract NAS3-25982 with 
Science Applications International Corporation. 



References 

1 . Sacksteder, K. R.: Twenty-Third Symposium (International) on Combustion, P. 1589, 
The Combustion Institute, 1990. 

2. Edelman, R. B. and Bahadori, M. Y.: Acta Astronaurica 13, 681 (1986); for previous studies of 
microgravity laminar diffusion flames, see references cited in this work. 

3 . Bahadori, M. Y., Stocker, D. P., and Edelman, R. B.: Effects of Pressure on Microgravity 
Hydrocarbon Diffusion Flames, Paper AIAA-90-0651, AIAA 28th Aerospace Sciences Meeting 
Reno, Nevada, January 1990. 

4. Bahadori, M. Y., Edelman, R. B., Stocker, D. P., and Olson, S. L.: AIAA J. 28, 236 
(1990). 

5. Bahadori, M. Y., Edelman, R. B., Stocker, D. P., Sotos, R. G., and Vaughan, D. F.: Effects 
of Oxygen Concentration on Radiative Loss From Normal-Gravity and Microgravity Methane 
Diffusion Flames, Paper AIAA-92-0243, AIAA 30th Aerospace Sciences Meeting, 

Reno, Nevada, January, 1992. 

6. Hottel, H. C. and Hawthorne, W. R.: Third Sympo sium on Combustion , pp. 254-266, 
Williams and Wilkins Co., Baltimore, 1949. 

7. Wohl, K., Gazley, C, and Kapp, N.: Third Symposium on Combustion, pp. 288-300, 
Williams and Wilkins Co., Baltimore, 1949. 



> •= 



o 
o 
to 



I* 

o 01 



o -o 

o u. 





>- t— 




►- X 


. 


, ►- o 




> -> 




< LU 


>> 


ec z 


•*-* 


c: 


o.*>- 


O UJ 


— > 


s$ 


4-i n 


1 u 


■— _J 


"w CT> v» 


X u. 


« O II 




v> i. E 


V 


o u « 


\ 






«_> E *- 





en in 
o u 
>- E 




J3 

E 



oi 



e»o l 



c — > 



eg.s 



■?£■ 13 t3 .2 



e J? o « © 
p "O p u a 

6 o .5 =6 1 

23E.SS 



(WD) m6j3H W»IJ 






Section 18 

'EFFECTS OF BUOYANCY ON LAMINAR GAS-JET DIFFUSION FLAMES* 

-Science Requirements Document- 
(Space Experiment); 



M. Y. Bahadori and R. B. Edelman 



October 1991 



EFFECTS OF BUOYANCY ON LAMINAR 
GAS JET DIFFUSION FLAMES 



- Science Requirements Document - 
(Space Experiment) 



M. Yousef Bahadori and Raymond B. Edelman 



SCIENCE APPLICATIONS INTERNATIONAL CORPORATION (SAIC) 

Thermal Sciences Division 

21151 Western Avenue 
Torrance, California 90501 




An Employee-Owned Company 



Submitted to: 

NASA LEWIS RESEARCH CENTER 
Cleveland, Ohio 44135 



October 18, 1991 



SUMMARY 

The proposed research is aimed toward the fundamental understanding of the behavior and 
characteristics of laminar gas-jet diffusion flames under micTogravity conditions. By eliminating 
buoyancy, fundamental scientific information can be obtained in relation to the structure and 
physico-chemical aspects of flames in both normal-gravity and reduced-gravity environments. 

The ground-based portion of the current program has resulted, to date, in some new and 
unique information, significantly enhancing our understanding of laminar diffusion flames in 
microgravity environments. However, unexpected characteristics have also been observed raising 
new questions on transport processes in diffusion flames in the absence of buoyancy. 
Observations and associated data show that earthbound facilities do not provide enough time to 
fully characterize the coupled physical and chemical processes which control the behavior of flames 
under microgravity conditions. 

This document presents the science requirements for a highly productive program of space- 
based experiments for studying laminar diffusion flames. Critical, fundamental questions will be 
addressed through the efficient combination of longer microgravity test times, advanced 
diagnostics, and comprehensive numerical modeling. 



TABLE OF CONTENTS 



Page 

I . Introduction 1 

II. Background 1 

III. Fundamental Characteristics of Laminar Diffusion Flames in Microgravity 10 

(a) Flame Characteristics 12 

(b) Ignition, Flame Development, and Extinction 13 

(c) Effects of Pressure and Oxygen Concentration 15 

(d) Radiation Measurements 27 

(e) Temperature Measurements 34 

(f) Predictions 37 

(g) Summary of Observations and Conclusions 41 

IV. Objectives 44 

V. Hypotheses and Data Requirements 45 

VI. Justification for Longer-Duration Experiments 50 

VII. Identification of Experiments 52 

Vni. Apparatus, Conditions, and Experiment Timeline 54 

IX. Relationship Between Data and Modeling 54 

X. Data Analysis 54 

References 61 



-u- 



I. INTRODUCTION 

Although diffusion flames have been the subject of research for many years they are not 
well understood and, while the effects of gravity on the burning process have been observed, the 
basic mechanisms responsible for these changes have yet to be determined. The goal of this 
research is to develop an improved understanding of gas jet diffusion flames. The laminar gas jet 
diffusion flame is selected for study because it embodies mechanisms important in all diffusion 
flames such as fires and practical combustion systems. Furthermore, due to its relative simplicity 
and ease of control, it lends itself to direct theoretical and experimental studies. However, existing 
theoretical models are deficient and improved models are required because they are essential to the 
interpretation of data, providing a means of theoretically isolating mechanisms influencing the 
observed behavior. The complexity of these mechanisms and their coupling establishes the 
requirement for data with which to verify assumptions inherent in the theoretical analyses. If 
gravity is significantly reduced, then its effect can be isolated, and the complication of buoyancy- 
induced convection would be removed from the problem. The combination of microgravity and 
normal-gravity data would provide the information, both theoretical and experimental, to improve 
our understanding of diffusion flames in general and the effects of gravity on the burning process 
in particular. Although earthbound experiments under reduced-gravity conditions have shown 
significant differences in flame behavior relative to normal-gravity conditions, the data is 
incomplete. This is mainly due to the limited available time in ground-based faculties, in addition 
to the limitations imposed by these facilities on the capability to conduct advanced diagnostics 
measurements. The existing data is confined to semi-quantitative visual observations on flames, 
and quantitative measurements (radiation, temperature, and species) from experiments currently 
being conducted in the 5.18-Second Zero-Gravity Facility and the upcoming KC-135 tests. This 
document describes the requirements of an experiment that if performed in space, would help to 
eliminate some major deficiencies in our understanding of gas jet diffusion flames. 

II. BACKGROUND 

The problem of fire safety has received the attention of researchers for many years which 
has resulted in a continuous effort to understand the complicated processes occurring in fires. 
Recently, interest in the fire problem has been expanded to include aerospace applications due to 
the concern over fires aboard spacecraft. The primary conclusion drawn from the existing base of 
information on fires is that a more fundamental understanding of diffusion flames is needed in 
order to predict their behavior and to control them. Although the significant effects of gravity on 
the burning process have been observed, the mechanisms responsible for these changes have yet to 



be determined A critical limitation of the existing earthbound experimental data is that only limited 
elapsed time is available to fully characterize the flame behavior under reduced-gravity conditions. 

Diffusion flames involve convective and diffusive effects and fires are a complicated type 
of this class of combustion process. Regardless of the type of fuel involved (i.e., liquid, solid, or 
gaseous), the basic mechanisms which control the burning process are the same. These include the 
processes of mixing, radiation, chemical kinetics, soot formation, diffusion, and buoyancy- 
induced convection. The laminar gas-jet diffusion flame embodies these mechanisms (see Figure 
1) while representing a basic element of more complex, convectively-induced turbulent flames (see 
Figure 2). 

In order to understand the mechanisms controlling a fire, laboratory-scale turbulent gas-jet 
diffusion flames have been extensively studied. These small-scale flames provide important 
information in relation to the processes occurring in practical combustion systems as well. 
However, turbulent flames are not fully understood and their understanding requires more 
fundamental studies of laminar diffusion flames in which, under normal-gravity conditions, the 
complication of convection-induced turbulence is eliminated and the interplay of transport 
phenomena and chemical kinetics becomes more tractable. But even this basic, relatively simple 
flame is not completely characterized in relation to soot formation, radiation, diffusion, and 
kinetics. Therefore, gaining an understanding of laminar flames is essential to the understanding 
of fires in which the same basic phenomena occur. By working under micrograviry conditions, the 
effects of buoyancy can be isolated and the results would form a basis to address, in logical order, 
the fundamental processes occurring in diffusion flames [1]. In the remainder of this section, we 
describe a cross-section of past experimental and theoretical efforts in relation to laminar diffusion 
flames under both normal-gravity and reduced-gravity conditions. 

Laminar gas-jet diffusion flames have been a subject of extensive analytical, numerical, and 
experimental research since the appearance of the classical work of Burke and Schumann [2]. 
Most of the analytical studies to date deal with the mathematical representations of flame shapes 
based on the Burke-Schumann methodology which neglects some of the phenomena critical to 
understanding the structure of these flames. These include transient effects, soot formation, non- 
uniform flow parameters, variable transport properties, coupling between the conservation 
equations, and finite-rate chemical kinetics. In all of the analytical approaches, the effect of 
buoyancy had to be neglected in order to enable one to obtain a closed-form mathematical represen- 
tation of flame shape, mass-fraction distribution and temperature distribution. Applications of this 
theory to laboratory-scale gas-jet diffusion flames [2-6] and to the flamelets of composite solid 



io««t-«w»* 




Fig. 1. Schematic diagram of a laminar gas-jet diffusion flame. The photograph shows a methane flame under both 
normaJ-graviiy and microgravity conditions. The extensive yellow and red in the background for normal- 
gravity flame is reflection from the chamber wall. Reproduction has increased the color intensity. 



Liminn 
lljmes 



Tunjwon 



fully dtvtlODid 
luOultrii llimis 




Volume flow rate (op Reynolds number) 



Fig. 2. Change in the flame height and behavior with increase in volume flow rate for a typical gas jet diffusion 
flame in normal gravity. 



propellants [7] have shown satisfactory agreement between the theoretical and experimental flame 
characteristics. Table 1(a) describes a selected number of analytical works performed on laminar 
gas-jet diffusion flames. Few numerical investigations of these types of flames under normal- 
gravity and microgravity conditions have been attempted. The results of numerical calculations of 
the set of conservation equations for normal-gravity [8] and zero-gravity flames [9-12] have shown 
that improved understanding is needed in relation to the coupled physico-chemical phenomena 
occurring under these conditions. Table 1(b) shows a selected number of numerical modeling 
approaches. Experimental investigations of normal-gravity flames have provided a large amount of 
information which is not quite obtainable from the theoretical studies cited above, i.e., flame 
structure, soot mass fraction and number density, radiative characteristics, and temperature, 
species and velocity distributions; see, e.g., [8-13]. 

Efforts to isolate the buoyancy effects in laminar flames have been limited, to-date, to 
experiments in the 2.2-Second Drop Tower [9-11, 14-19] and 5.18-Second Zero-Gravity Facility 
of NASA-Lewis Research Center [20, 21], and accompanying numerical modeling [6, 9-1 1]. The 
2.2-Second Drop-Tower experiments involve a self-contained gas-jet diffusion flame apparatus 
which is allowed to fall inside a falling drag shield. The apparatus is dropped (after establishing 
and recording the flame in normal gravity [9, 10, 14-17] or igniting the flame while in micTogravity 
[11, 18, 19]) and the flame is observed throughout the duration of the fall. This provides time- 
resolved data on the development of the flame shape for the near- zero-gravity condition achieved 
during the drop. In the 5.18-Second Zero-Gravity Facility, the experiment package falls inside an 
evacuated tube, and the flame is ignited while the package is falling [20, 21]. 

Attempts have also been made to study the gas-jet diffusion flames under elevated-gravity 
conditions achieved in a centrifuge [22] and under negative-gravity conditions by inverting the 
flame using a downward-flow configuration [1, 17, 23]. Studies have been conducted on soot 
formation from laminar diffusion flames under normal-gravity conditions in the pressure range of 
1.0-10 atm [e.g., 13]; there is no study to-date which quantitatively determines the effect of 
pressure on soot formation in the absence of buoyancy. Experimental studies have been underway 
on the behavior of Burke- Schumann type diffusion flames in microgravity environment [e.g., 24]. 
Results of selected experimental works on laminar flames are presented in Table 1(c). We sec that 
while the effects of buoyancy are apparent in all of the studies presented in Tables 1(a)- 1(c), it has 
not been strictly isolated (except for the flames of drop-tower experiments), and a large gap in the 
data base exists in this context. 



Table 1(a). Selected Analytical Studies on Laminar Diffusion Flames 



AUTHOR! S) 


KfTHCD 


CHARACTERISTICS 


DATA USED 


DEFICIENCIES 


COMMENTS 


Burke tnd 


Analytical represent- 


Assumptions of equal and 


Methane, dty gas. 


No eccount of the 


Theoretical 


Schumann 


ations of the shapes of 


constant velocities. 


and CO flames are 


formation of pro- 


flame heights 


(1928) ; 


laminar, cylindrical ly- 


transport coefficients. 


used. Air flow 


ducts of combust- 


are compared 


Ref. [2] 


symmetrtc, co-axial gas- 


and physical properties 


rate*6-42 ft'/hr. 


tlon; leek of 


with the ex- 




jet diffusion flames 


of the two gas streams; 


fuel flow rate* 
2-25 *r 3 /hr. 


finite-rate kin- 


perimental 




under steady-state 


flaae-sheet approxi- 


etics and soot- 


results 




conditions. 


mation; Lewis number of 


Insldt radius* 


formation mech- 


under norma 1- 






unity (which decouples 


1/4-5/16 In. 


anism; no effects 


g; qualita- 






the energy and species- 


Outside radius* 


of radiation; no 


tive agreement 






conservation equations); 


1/2-5/8 In. 


axial diffusion; 


is obtained 






constant pressure (which 




assumption of de- 


between ex- 






eliminates the momentum 




coupled conservat- 


periment and 






equation); absence of 




ion equations; no 


theory. 






buoyancy; no axial 




transient effects. 








diffusion. 








Astavin and 


Same as Burke and 


Same as Burke and 


Parametric values 


Same as Burke and 


Families of 


Ryezantsev 


Schwann (1928) but for 


Schumann (1928). 


of velocities. 


Schumann (1928). 


flame shapes 


(1979) ; 


fuel and oxidizer sep- 




diffusion coef- 




and temper- 


Ref. [3] 


arated by a flow of an 




ficients and tube 




ature contour; 




Inert gas (a) from each 




diameters are used 




are obtained. 




other and (b) from the 




for methane/air 




No comparison 




wall for both cylindrical 




flames. 




with data. 




and rectangular burners. 








- 


Penner , 


Same as Ourke and 


Same as Burke and 


The data of Burke 


Same as Burke and 


The results 


Bahadori and 


Schumann (1928). 


Schumann (1928) but for 


and Schumann (1928) 


Schumann (1928) 


are 1n good 


Kennedy (1984) 




constant arbitrary flow 


and Mitchell et al. 


except for the 


agreement 
with experi- 


Ref. [5] 




velocities, transport 


(1980) are used. 


effect of axial 






parameters and physical 




diffusion. 


mental data. 






properties in the pre- 












sence of axial diffusion 












with unimolecular decom- 












position reactions of 












fuel and oxidizer. 








Beckstead 


Analytical representation 


Same as Burke and 


Typical coeffic- 


Same as Burke 


Flame stand- 


(1981) ; 

Ref. [7] 


of the heights of 


Schumann (1928) but in 


ients of diffusion 


and Schumann 


off distances 


laminar, cylindrleeny- 


the presence of axial 


and velocities for 


(1928) except for 


are obtained. 




syirmetrlc flamelets of 


diffusion. 


evaporating solid 


the effects of 


No data avail- 




composite solid propel- 




propellent species 


axial diffusion. 


able. 




lents under steady-state 




ere used. 








conditions. 










Bahadori 


Analytical representation 


Same as Burke and 


The data of Cochrar 


Seme ts Burke and 


Flame devel- 


(1990); 
Ref. [6] 


of the shapes of laminar. 


Schumann (1928) but 1n 


(1972) and Cochran 


Schumann (1928) 


opment with 


cylindrical ly-synmetrie, 


the presence of axial 


and Maslce (1970) 


except for the 


time Is ob- 




transient, co-axial gas- 


diffusion. 


are used, e.g.. 


effects of axial 


tained. The 




jet diffusion flames. 




nozzle radius -0.186 


diffusion end 


results In- 








cm, methane velo- 


tram lent pheno- 


dicate that 








dty*2S.B cza/sec. 


mena. 


times greater 








etc. 




than S.o sec 
»rt needed for 
flame to reach 
near-steady 
state under 
zero-g 
conditions. 



PeVGt-2, 



Table 1(b). Selected Numerical Studies on Laminar Diffusion Flames 



AUTHOR(s) 


METHOD 


CHARACTERISTICS 


DATA USED 


DEFICIENCIES 


comtNTs 


Mitchell, 
Sarofim and 
Clomburg 
(1980) ; 
Ref. [8] 


Numerical solution of 
governing conservation 
equations for mass, 
momentum, species, and 
energy, using t global 
chemical reaction tt 
the flame boundary with 
infinite rate. A steady- 
state normal-g methane/ 
air flame was studied, 
which was laminar, 
confined, and txisy- 
mmetric. 


Flame-sheet concept for 
locating the stoichio- 
metric fuel-oxygen 
Interface and hence, the 
points of heat release. 
Natural convection 
effects tnd variable 
thermodynamic tnd trans- 
port properties are 
allowed. Both axial 
and radial convection 
and diffusion are In- 
cluded. 


Inner tube dlt- 
meter-1.27 cm; 
outer tube dit- 
meter'5.18 cm, 
methtnt flow rite* 
5.7 cm 3 /sec; tir 
flow rtte»187.7 
cm 3 /sec. 


No detailed 

chemical kin- 
etics, soot 
formation or 
radiation. 


Temperatures, 
species con- 
centrations 
and velocit- 
ies are ob- 
tained. The 
results show 
the Important 
effect of 
buoyancy, by 
which, the 
velocity In- 
creases by a 
factor of 40 
inside an 
overventllat- 
ed CH4-a1r 
flame. Good 
comparison 
with experi- 
mental data 
1s obtained. 


Klajn and 
Oppenheim 
(1982) ; 
Ref. [12] 


The numerical model 
treats the effects of 
exothermlclty on the 
contours of gaseous, 
unconflned Jet diffusion 
flames with negl iglble 
buoyancy effects. 


Simplifying assumptions 
of flame-sheet approxi- 
mation, Schmidt and 
Prandtl numbers of unity 
and constant specific 
heats. Solutions *rt ob- 
tained in closed alge- 
braic form. Axial 
diffusion is not included 


Nondimensiontliied 
values of diameters, 
stoichiometric 
ratios, velocities , 
etc. are used 1n 
a parametric form. 


Lewis and Prandtl 
numbers of unity; 
Infinite-rate 
glohal chemical 
reaction; no rad- 
iation or soot 
formation. 


Fair agree- 
ment when 
compared with 
the experi- 
mental data 
under zero-g 
conditions 
of Cochran 
(1972) and 
Haggard tnd 
Cochran 
(1973). 


Edelman, 
Fortune, 
Wellersteln, 
Cochran, and 
Haggard (1973) 

Ref. [10] 


Numerical solution of 
conservation equations. 
It is shown that kinetic 
effects are of pritiary 
importance in flames 
under reduced gravity. 
Other controlling fact- 
ors ire t«itl diffusion, 
radiation and transient 
effects. 


The model includes the 
coupled effects of 
Inertia, viscosity, 
diffusion, gravity, tnd 
combustion in t boundary- 
layer type formulation. 
Axial tnd radial con- 
vection and radial dif- 
fusion are considered. 


Comparison with 
drop-tower dttt of 
Cochrtn tnd Htslct 
(1970) tnd Cochran 
(1972); see 
Table 1(c). 


No detailed kin- 
etics, soot form- 
ation, ailtl dif- 
fusion, and rtd- 
Ittion. 


Good agree- 
ment for 
normal-g 
flames and 
fair agree- 
ment for 
steady-state, 
low-Re, zero- 
g flames of 
NASA- Lew is 
drop-tower 
experiments. 


BahadoH , 
Edelman, 
Stoeker, tnd 
Olson (1990); 
Ref. [11] 


The model of Edelman, 
tt al. (1973). 


Modified radiation 
model was used to 
accommodate the effects 
of gas-phase radiation. 


Comparisons with 
drop-tower dtta 
of Cochran (1972) 
and Haggard and 
Cochran (1973), 
1n addition to 
the new data . 
Also, comparisons 
with the predic- 
tions of Klajn 
and Oppenheim 
(19«2). 


Absence of radia- 
tion from soot; 
axial diffusion; 
soot processes. 


Good tgree- 

ment for lero 

grtvlty 

flames, tnd 

excellent 

tgreement 

with the 

predictions 

of Kltjn and 

Oopenhelm 

(1982). 



\S 



$foo*^ 



Table 1(c). Selected Experimental Studies on Laminar Diffusion Flames, 



AUTHOR (s) 



Mitchell , 
Sarof im 
and Clanburg 
(1980) ; 
Ref. [8] 



METHOO 



Confined, axlsymnetrlc, 

laminar methane-air 
diffusion flames. 



CHARACTERISTICS 



Flower and 
Bowman (1983) 
Ref. [13] 



Structures of 2-0. lam- 
inar ethylene-alr run 
under normal-g and 
elevated-pressure cond- 
itions trt studied. 



Distributions of temper- 
ature, velocity and 
species were measured. 
The effects of buoyancy 
have not been isolated. 



DATA USED 



Cochran and 
Maslca (1970) 
Ref. [H] 



Cochran 
097?) ; 
Ref. [15] 



NASA/Lewis drop-tower 
experiments (see the 
text for details). 



Soot aiass fraction and 
number density is ob- 
tained. No significant 
effect of pressure on 
the slxe of soot parti- 
cles 1s observed. 



Inner tube dla.* 
1.27 on; outer 
tube d1a.-5.l8 en; 
■ethane flow rate* 
5.7 c-i 3 /see, air 
flow rate*187.7 
e»3/see . 



DEFICIENCIES 



COMMENTS 



No kinetics, soot 
formation or rad- 
iation effects. 



Pressure range* 
1.0 - 2.5 (tie. 
A rectangular 
burner was used 
with air velocity 
• 22 cm/sec and 
fuel velocity* 
7 cat/sec. 



Only gas temper- 
ature and soot 
data are obtained 



Importance of 
buoyancy Is 
demonstrated, 
by which, the 
velocity in- 
creases by a 
factor of 40 
Inside an 
overventllatet 
CH4-a1r flame 
The data are 
In good agree- 
ment with the 
results of 
model (ng. 



Flame behavior 1s ob- 
served during the 2.2- 
sec drop of the test 
chamber. Quiescent air 
for all of the drop- 
tower experiments ex- 
cept those of Haggard 
(1981) which were stud- 
ied for Inverted flames 
(under normal-g cond- 
itions) end zero-g 
flames In the presence 
of fore*d-a1r velocities 
see the text for details 



ChValr flames; 
nozzle radius* 
0.186 - 0.442 cm; 
methane flow rate* 
1.2 • 5.3 «3/see. 



CH 4 /a1r flames; 
nozzle radius* 
0.051 - 0.113 
em; fuel flow 
r«te-0.7 - 1J.2 
e»3/»tc. 



Only flame 
lengths and be- 
haviors are ob- 
served . No 
measurements on 
quantitative 
data was possible 
due to the short 
periods of drop. 



Increase in 
soot yield 
(mass of 
soot/mass of 
fuel) as the 
pressure to 
the power 
0.5 - 1.0 
was observed 
Maximum and 
Integrated 
soot volume 
fractions 
Increased 
as the pres- 
sure to a 
power be- 
tween 1.5- 
2.0. 



Sudden Incre 
ase In gravity 
level result- 
ed In Immedi- 
ate reduction 
In flame len- 
gth followed 
by flame ex- 
panding away 
from the 
burner and 
final exting- 
uishment. 



Steady-state, 

transient and 
extinguished 
flames exist 
ed In zero 
gravity. 
Flames were 
- SOt long- 
er and wider 
In zero-g 
than In 
normal-g. 



srsss&s 



OF POOR 



Table 1(c). Cont. 



AUTHOR(s) 



METHOD 



CHARACTERISTICS 



DATA USEO 



DEFICIENCIES 



COfttENTS 



Haggard and 
Cochran 
(1972) ; 
Ref. [16] 



Haggard 
(1981); 
Ref. [17] 



Ethylene end pro- 
pylene diffusion 
flanes; nozzle 
rad1us*0.O51 - 
0.083 ca; fuel flow 
rete*0.4 - 3.6 
ca 3 /see. 



CH«/e1r 

fuel-no 

0.05 

coaxial 

eir-noz 

1.4 ca 

plane; 

rete«l 

ca 3 /see 

rete^O? 



flues, 
zzle radius* 
0.30 ca; 

convergent 
tie rtdlus* 
at exit 
fuel flow 
SS • 10.3 

•1r flow 
597 eaVsee. 



Flame length 
Is described 
1n terms of 
flow para- 
■eters end 
fuel prop- 
erties. 
SUble zero- 
9 lengths 
vere observed. 



Experimental 
end coaputed 
flaae lengths 
for normal -g 
have shown 
good igree- 
aent. flame 
extinguish- 
es rt upon 
entry Into 
tero-g was 
studied. 
Relatively 
low forced- 
i1r velocit- 
ies (>10ca/ 
sec) were 
sufficient 
to sustain 
coabustlon 
1n zero 
gravity. 



AHenkrich. 
Eichhorn, Hsu, 
Brine fc, and 
Cera 11 os 
(1976); 
Ref. [22] 



Gas-jet diffusion flames 
under elevated gravity 
conditions achieved In a 
centrifuge. 



Hydrogen, aethane, 
ethane and propane are 
used. 



1.83-a dlaaeter 
centrifuge; burner 
Inside dlaaeter* 
0.05 • 0.21 ca; 
rotational speed* 
200 rpa. 



Only photographs 
are obtained and 
flaae lengths are 
■easurtd. 



The results 
Indicate 
that flame 
length and 
carbon lum- 
inosity de- 
crease with 
Increasing 
buoyancy. 
Alto, flaae 
lift-off 
and exting- 
uishment 
■ere obser- 
ved as the 
g-level 
increased. 



dmure and 
Ukawa (1961); 
Ref. [23] 



Gas-Jet diffusion flaaes 
under negatlve-g condi- 
tions by Inverting the 
flaae using a downward- 
flow configuration. 



Data are obtained for 
both nonaal-g and in- 
verted fla 



City gas flaaes 
are used. Fuel 
nozzle dla.'l.S, 3 
and 6 aa; air duct 
d1a.*52 and 18 aa; 
air flew rate S 
10 a/sec, fuel flow 
rate £ 40 a/sec. 



Only flaae 
lengths are ob- 
served . 



Flaae length 
is proport- 
ional to the 
fuel flow 
rate and in- 
versely pro- 
portional to 
the diffus- 
ion coeff- 
icient. 



ff?3£#£ 



Table 1(c). Cont. 



AUTHOR(s) 


METHOD 


CHARACTERISTICS 


OATA USED 


DEFICIENCIES 


COMMENTS 


Edelman 


A downward-flow configur- 


Temperature, flame shape, 


Air Inlet velocity 


No quantitative 


Importance 


and 


ation Is used to study 


etc. are obtained. Stag- 


•6.4 ft/sec; met- 


Information on 


of recircul- 


lanadoM 


coaxial gas-jet diffusior 


nation points ere pre- 


hane Injection vel- 


soot, velocity. 


ation In 


(1986); 


flames. 


dicted Indicating that 


odty0.17 - 1.2 


species, radiation 


Inverted 


Ref. [1] 




unattached recirculat- 


ft/sec; fuel -tube 


etc. 


flames Is 






ion zones exist above 


diaaeter-0.19 1n.; 




demonstrated. 






the centerllne down- 


duct diameter«6.0 




The tip of 






stream of these flames. 


In. 




the flame 
spread out 
and the 
flame was 
capped with 
a concave 
upward 
surface. 


Stocker 


Co-axial Jets of hydro- 


Flames were Ignited 1n 


Fuel -tube diameter 


No quantitative 


Shape and 


(1990); 


carbon fuels burning In 


microgravlty. Both 


■0.56 cm, outer 


Information on 


size of flame 


Ref. [24] 


air using a Burke- 


equal and non-equal 


d1a. > 4.66 cm. 


soot, temperature 


1s Strongly 




Schumann type configu- 


reactant velocities are 


Methane, ethane, 


velocity distri- 


Influenced 




ration are studied In 


used for methane. 


and propane were 


bution, radiation 


by gravity. 




2.2 seconds of micro- 


Ethane and propane 


used. Velocities 


etc. 


Open tip was 




gravity. 


flames had non-equal 
fuel and air velocities. 


up to 15 cm/s. 




observed for 
ethane and 
propane 
flames. 


Bahadorl 


2.2-sec. microgravlty 


Flames were Ignited 1n 


Nozzle radius ■ 


No quantitative 


Flames with 


and 


experiments conducted 


microgravlty. Increase 


0.048 and 0.074 


Information on 


181 Oj were 


Stocker 


for methane and propane 


1n oxygen concentration 


cm. Flow rates ■ 


soot, radiation. 


entirely 


(1989); 
Ref. [IS] 


flames with Oj-N? . 
mixtures containing 


reduced the flame 


1.75 and 5.25 


temperature, etc. 


blue In 


height. Underventl la- 


cc/s for methane 




microgravlty 




181 to 305 oxygen 1n 


ted behavior was obser- 


and 0.64 and 0.96 




but not 1n 




quiescent environments 


ved for propane flames. 


ce/s for propane. 




normal gra- 




at 1 atm. 








vity. 


Bahadorl , 


Ignition and behavior of 


Flames were Ignited In 


Nozzle Inside rad. 


No quantitative 


Flames previ- 


Edelman, 


methane and propane 


microgravlty. Blue CH4 


- 0.051 and 0.0825 


measurement of 


ously repor- 


Stocker, 


flames burning 1n air 


flames were observed at 


cm. Flow rate • 


temperature , 


ted as extin- 


and Olson 


are studied In 2.2 s 


low flow rates. Flame 


1.0-3.0 ce/s for 


radiation, etc. 


guished were 
burning dur- 


(1990); 


of microgravlty. 


heights were 1n good 


methane and 0.5- 




Ref. [11] 




agreement with predic- 


1.5 ec/s for 




ing the 2-0 s 






tions. Open tips were 


propane . 




of micro-g 






observed 1n mlcro-g. 






due to 

unproved 

photography. 


Bahadorl , 


Propane flames burning 


Flames were ignited 1n 


P ■ 0.5, 1.0 and 


No quantitative 


Observed open 


Stocker, 


In quiescent air Mere 


microgravlty. Flames 


1.5 atm. Fuel 


measurements of 


tip and soot 


and Edelman 


studied in 2.2 sic, of 


show non-aonotonlc 


flow rate • 0.5 - 


temperature, 


escape. Soot- 


(1990); 
Ref. [19] 


microgravlty at 


height variation 1n 


1.5 cc/s. Tube 


radiation, etc. 


ing was Incr- 


different pressures. 


microgravlty. 


diameter ■ 0.15 cm. 




eased with 












pressure 












Increase. 


Bahadorl , 


Methane and propane 


Radiation aeasureaents 


Nozzle radius • 


No other quantita- 


Radiance was 


Edelman, 


flutes were studied In 


using a thermopile 


0.0825 cm. CH 4 
flow rate ■ 1 .0 - 


tive measurement 


uo to 10 


Sotos, and 


quiescent air at 1 eta 


detector. Flaaes were 


except flame 


times larger 


Stocker 


pressure 1n 5. IB-sec. 


Ignited In microgravlty. 


3.0 ec/s; C 3 H S 


radiation. 


1n aicro-g. 


(1991); 


of microgravlty. 


S sec. was not sufficient 


flow rate - 






Ref. [20] 




for radiation to reach 
Steady state. 


0.5 • 1.5 cc/s. 






Bahadorl , 


Methane and propane 


Flames were Ignited In 


Nozzle radius ■ 


Only temperature 


Temperatures 


Edelman, 


flames studied In quies- 


m1cro-g. Temperature 


0.0825 cm. CHi 
flow rate ■ 1.0 - 


data are reported. 


are lower 


Sotos, and 


cent air at 1 atm In 


measurements using a 




compared to 


Stocker 


5.16 sec of mlcro-g. 


rake of 9 thermocouples. 


3.0 cc/s; propane 




normal -g. 


(1995); 
Ref. [21] 






flow rate - 0.5 - 




5 sec was not 






1.5 cc/s. 




sufficient 












to reach 








1 




steady state. 



srsss&s 



III. FUNDAMENTAL CHARACTERISTICS OF LAMINAR DIFFUSION 

FLAMES IN MICROGRAVITY 

In this section, we present some of the important findings of the ground-based portion of 
the current program. Both experimental and theoretical aspects of this effort are described. The 
new findings, when analyzed and discussed, provide a set of objectives and associated justification 
for conducting a space-based experiment; these are presented in the later sections. 

The experimental portion of the current ground-based program was designed to provide 
both qualitative and quantitative measurements in the chamber shown in Figure 3. These data are 
being obtained by the combined use of the 2.2-Second Drop Tower, 5.18-Second Zero-Graviry 
Facility and the KC-135 aircraft. To-date, quantitative data have been obtained during the 
approximately 5 seconds of low gravity in the form of temperature and radiation measurements. In 
addition, species concentrations will also be measured. Direct observations using cinematography 
have been obtained, revealing additional flame-visualization information in relation to flame color, 
luminosity, and dimensions. 

The current theoretical work consists of the development and upgrading of comprehensive 
numerical models that are designed to predict flame characteristics, including temperature, velocity, 
and species fields, radiation from particulates and combustion products, and flame shapes for 
arbitrary gravity levels, in the presence of soot formation and oxidation, thermophore tic effects, 
finite-rate kinetics, and multi-component diffusion. 

In the current study, the effects of pressure, oxygen concentration, fuel type, nozzle size, 
and fuel-flow rate (i.e., Reynolds number) have been examined. A comprehensive set of data 
(both qualitative and quantitative) have been obtained. Oxygen-nitrogen mixtures in the range of 
15% to 50% oxygen at chamber pressures ranging from 0.5 atm to 1.5 atm have been used as the 
oxidizing environment Both methane and propane flames have been studied in these quiescent 
environments under normal-gravity and microgravity (both 2.2-sccond and 5.18-second) 
conditions. Nozzle sizes of 0.051 to 0.0825 cm (inner radius) have been used. The fuel-flow 
rates (when convened to standard conditions of 1 atm and 298 K) were 0.5 - 1.5 cm 3 /sec for 
propane and 1.0-3.0 cm 3 /sec for methane. 

Results of the research conducted to date under the current program have shown important, 
and in some cases unexpected phenomena. Some of these findings are presented below in the 
form of results, issues, and their impact on the direction of this research. 



10 



ruu Klin 

THERMOCOUPLES 



RACIOMETER 




5PWU ELECTRODE 
GAS SAWLIHG PROBES 



> 



tern CAKEItA 



RAO! METER 



*■■!"" J ^ GAS SAMPLING PROBES 



AMO therpocoup LES 




Wtlt CJM.HA 



SPAM ELECTRODE 



FUEL N0Z2U 



Fig. v Experiment package for ihc 5.18-Sccond Zcro-Graviiy Facility and KC-135 tests; current microgravity 
laminar diffusion-flame program. 




1-g minimum 



1-g maximum 



Microgravity 



Fig. 4. Normal-gravity (both minimum and maximum heights of flicker) and microgravity flames of propane 
burning in quiescent air at 1 atm; fuel-flow rate * 1.0 cm 3 /sec. The extensive yellow and red in the 
background for normal-gravity flame is reflection from the chamber wall. Reproduction has increased the 
color intensity. 



11 



(a) Flame Characteristics 

Laminar diffusion flames of hydrocarbons under microgravity conditions have shown 
distinct characteristics relative to normal-gravity flames [1]. Compared to the flames in normal- 
gravity environments, larger, sootier, and somewhat globular flames are observed in microgravity. 
This is due to the significant reduction in the buoyant force, which makes diffusion the dominant 
mechanism of transport. As a result, increased residence time, enhanced soot formation, radiation 
cooling due to the larger flame size, and the possible onset of a chemical-kinetics limitation on the 
heat-release process are apparently responsible for the very different characteristics of these flames 
compared to those in normal gravity. In normal-gravity environments, laminar diffusion flames of 
hydrocarbons flicker and are generally yellow, whereas their microgravity counterparts are flicker- 
free, with colors ranging from orange to red to entirely blue, depending on the oxidizer 
composition and pressure. 

In gas-jet diffusion flames, the heat-release mechanism is not uniform throughout the flow 
field. Pyrolysis, producing soot and partially oxidized species, dominates in the near-jet region, 
while continued heating and increased residence time result in the tendency to burn off these 
species downstream. However, pyrolysis and soot formation reduce the temperature level in the 
flame and, with the aid of the increased residence time, sooting is enhanced. This, in turn, results 
in enhanced radiation, more cooling, and delayed soot burn-off downstream of the flame. This 
effect is much more pronounced in microgravity flames compared to those in normal gravity, due 
to the lack of buoyancy. 

Figure 4 shows a typical normal-gravity and microgravity propane flame. The colors 
indicate the presence of particular species in hydrocarbon flames, e.g., blue-violet region (outer 
region at the base showing the spectrum of burning CO) and bluish-green region (showing C 2 and 
CH emissions, where C 2 bands appear green shaded toward violet and CH bands appear violet- 
blue). A highly luminous zone (bright yellow) indicates the thermal emission of burning carbon 
particles. This luminous zone becomes yellow, then orange, and then red and dark red toward the 
boundary of the visible region, as the temperature of the unbumed soot decreases. The reddish- 
orange color of the flame is due to the soot at cooler temperatures than would be expected for 
typical burn-off conditions. 

The color of the radiating soot, in conjunction with Wien's law, can be used as an indicator 
of the approximate local flame temperature. The cloud of burning soot particles is yellow, i.e., 
~0.58-^m wavelength (1370 K<^< 1670 K, where 1670 K is the temperature of a white hot 



12 



body). As the temperature decreases, the soot particles become orange red (-1370 K, -0.6 jim), 
then cherry red (-1170 K, 0.6-0.65 urn), and finally dull red (-970 K, 0.65-0.70 Jim). These 
effects are shown clearly in Figure 4, which indicate cooling by pyrolysis and enhanced radiation 
due to increased soot formation. 

Flame radiation is mainly due to the presence of soot particles, and the interaction between 
kinetics and radiation is directly influenced by soot generation and bum-off. These mechanisms 
are accentuated in the absence of buoyancy. The orange-reddish appearance of the microgravity 
flame in Figure 4 is an indication of pyrolysis in the hot fuel-rich portions of the flame which 
results in appreciable amounts of solid carbon. Studies of microgravity flames have shown that 
flame temperature is reduced due to the large globular flame surface and continuum radiation from 
soot. Thermal radiation can result in an energy loss of up to 40% of the heat input which augments 
the kinetic effects that can contribute to flame extinguishment [10]. The studies also suggest that 
pyrolysis and soot formation are of potential importance in terms of their effect upon the local 
molecular weight (buoyant force). 

(b) Ignition, Flame Development, and Extinction 

In previous studies [e.g., 14, 15], flames were ignited in normal gravity and then subjected 
to the microgravity condition. As a result, a transient period due to the sudden removal of the 
gravitational acceleration was imposed on the flame, which affected its characteristics and 
development toward steady state. Immediately after the release of the apparatus, the (steady-state, 
normal-gravity) flame height decreased and reached a minimum value in times on the order of 0.05 
sec; see Figure 5(a). The flame height then continuously increased leading to either (reported) 
extinguishment, (apparent) steady state, or a transient situation where the flame needed more time 
to show its final behavior. The transitional behavior is due to the accumulation of hot products of 
combustion attendant with the sudden loss of buoyancy-driven convective flow. Such an 
occurrence was believed to lead to quenching in some cases or possibly a slow readjustment to a 
steady- state microgravity flame (see Figure 5). 

The ignition approach previously used in the drop-tower experiments [14, 15] exposes an 
established normal- gravity flame to a sudden change in the gravity level. The reported micro- 
gravity flame behavior has been due to this step change in the gravity level. The normal-gravity 
ignition method also suggests that the microgravity flame behavior may be affected by the 
flickering normal-gravity flame at the instant the package is subjected to the microgravity condition. 
However, poor photography has resulted in erroneous conclusions on extinction of some of these 



13 



9 
C 

V 



3 3- 



30 - 



2 3 



2.0 



Near- 
steady- — Jk 
state • 
rero-g / 
length * 


._.,•.. ••,.«.. < 


- • 
• 


(a) 




• 






• 


Fuel Dm rjU 
• 2.1 cc/J 




— • 
• 


hull rutwt 
- 0.051 a 




* %y , 

-v; 

t Mi 
Normol T, 


Awt. normol-f 
tengrti 

n. nre-f fr^m 




1 1 1 


1 



-0 4 4 8 12 1.6 2.0 

Time, t 



2 9 - 



E 

V 



5. 2 3 

c 
« 



o 
u. 



2 1 



17 









Reported 


- 






extinction \ 
• • 




• 


% 


• 
• 

• 


- 


F«1 nam rite 
■ 2.1 ec/i 


• 


• 
• 




•t.IMn 




.* <w 


- 




.• 


• 




' 


1* 
1 • 

i- 


• 

Ztro-f 

— — 1 1 



■0 4 



4 

Tim«, s 



01 



Fig. 5. Flame length as a function of time in microgravity for a methane-air diffusion flame that (a) reached near- 
steady-state, and (b) was reportedly extinguished; reproduced from [15] and [14], respectively. 



o o o o o 



(a) 



CM «f J.» 



V***' 


■•lilt 

IHiailal 


'Ml Mm 


• 

■ 

o 

A 




0.«>1 

».wu% 

1 


j.a 
».o 
■•• 

J.O 
1.0 

i.i 



a-~r 



i ■ n ! i ■ t ' j ' i' i ' ^b 1 iU ' ix ' ii ' i » 



12- 
11 . 
10- 
• ■ 



o e • ° • ° 



(b) 






I •■ "• 



• ! • • 



btfrf t.l 



ll—l 




r-ii riM 


• 
• 
■ 
o 

A 



1 

•.an 

1 









TDM (*M>- 



04 CM 1.1 1J 2.0 2.4 U »J ».• 



Fig. 6. Flame heights for (a) methane, and (b) propane, following ignition in microgravity. Note that extinction 
was not observed for these flames. Those which do not show a complete height development with time 
could not provide a conclusive result in relation to the flame height due to a very faint flame tip; reproduced 
from [11]. 



14 



microgravity flames [14, 15]; see Figure 5(b). Recently, by improving the photographic 
technique, flames tested in the 5.18-Second Zero-Gravity Facility under identical conditions to 
those reported as extinguished, continued to burn throughout the available time of 5 seconds. As a 
matter of fact, none of the tests conducted to date under the current program have resulted in flame 
extinction in microgravity. In addition to the improved photography, these flames were ignited 
after the release of the package, therefore eliminating the complications arising from subjecting a 
normal-gravity flame to microgravity condition. 

Figure 6 shows the behavior of a set of methane and propane flames following ignition in 
microgravity. The data have shown that most of the flames do not reach steady state in the 2.2 
second available time. This prompted the use of the 5.18-Second Zero-Gravity Facility, the results 
of such tests will be discussed later. 

Figure 7 shows a frame-by-frame analysis of a propane flame ignited in microgravity [11]. 
Frame number 35 shows the flame during rapid deceleration as the experiment package is brought 
to rest at the end of the drop. The high level of induced convection causes blowoff within <0.1 
second. After the impact, re-ignition can be seen due to the spark pulses (in this case, the ignitor 
was active throughout the 3.0 seconds of experiment duration), resulting in normal-gravity 
behavior, i.e., flame flicker. 

In order to theoretically study the development and behavior of a flame following ignition 
in microgravity, a model was developed in the course of the current program for transient, laminar 
jet diffusion flames in zero gravity [6]. The model includes axial diffusion, and the solution yields 
the development of the flame boundary from ignition toward steady state, and reduces smoothly to 
the classical solution of laminar diffusion flames as, independently, axial diffusion becomes 
negligible, and steady state is approached. Figure 8 compares the theory with results of the 
experimental work reported elsewhere [14, 15], and shows satisfactory agreement for the increase 
in flame height with time. 

(c) Effects of Pressure and Oxygen Concentration 

In order to study the effects of oxygen concentration and pressure on micrograviry flames, 
tests have been conducted under various oxygen/pressure conditions. The results show that 
environmental conditions have a much more pronounced effect on flame characteristics, color, 
luminosity, and sooting behavior in micrograviry compared to normal gravity. Sooting was 
eliminated in microgravity hydrocarbon flames at 18% 02-0.5 atm, and the flames were entirely 



15 




s 

1 * 

> u 



E* 



•o « > a 

o c > — 

Ml: 



a j< ^ 



o qo 



S 



00 

o 

© 

II ' 

2 .2 
.2 .oo 

1 cq 



j> 8 5 _ 

I sal 

1* I 



£3 a 



•s 

u 



a 



s w 

5- S 

> s 

S ° 

E = 
o 

o Sr 

.S2 5 

- £ 



JE 
o 

>>© 
i - 

S B 



60 



•si*: 

1-1 . a so 

j 1 -HI] 

- 13 -s 

» c c « 

— C o <— 
U « 3 O 

o ..■do 

tew" 
1^3 | 

E u 

S •« 






. o 






E 
u 

II 

r 



u 

u 
in 



= ^ € -o 
13 — •> e 



c 
s 

5 



16 



w<£ 



\S 






%** 



6.0 



5.0 



■ 

-* 4.0 



I 3.0 



2.0 



(a) 



Reported 
Extinction 



h„ • 5.3 



■ 0.75 cm 2 /! 




Eiptrinvnt 
Trtnsitnt nodtl 
I—-— Th«or«t1c*l S.S. 



J_ 



*! • 0.1M em 
u • 25.76 emn 



0.5 



1.0 



1.5 



2.0 



2.5 



3.0 



Tli 




Fig. 8. Comparisons between the analytical [6] and experimental [14, 151 variations of flame height with time. 
The experimental work reports flame extinction (see a) at approximately 0.6 sec, and reaching an apparent 
steady state in 0.4 sec (see b). The binary diffusion coefficient of methane in nitrogen, D, has been used 
parametrically to show its effect. The values of 0.85 and 1.0cm 2 /s correspond, respectively, to 
temperatures of 650 and 750 K. Here, h„ is the theoretical steady state-flame height, u is the gas velocity, 
and R | is the inner (nozzle) radius. Axial diffusion has been included in the analytical solution: reproduced 
from [6]. 



17 



c f 



blue (see Figure 9), whereas their normal-gravity counterparts were yellow, luminous, and very 
similar to normal- gravity flames under atmospheric, or even high-pressure/high-oxygen- 
concentration conditions. This has a very important implication, namely, reduced radiation loss 
and reduced hazard of flame spread to surrounding combustible materials by reducing oxygen 
concentration and/or pressure in microgravity environments. In addition, there is a possibility that 
flammabiliry limits may be quite different in microgravity and normal-gravity flames under these 
reduced oxygen -concentration and pressure conditions. 

Figures 10 and 1 1 show, respectively, methane- and propane-flame heights as a function of 
oxygen concentration for different flow rates under both normal-gravity and microgravity 
conditions. The flames are taller and significantly wider in microgravity, the reason being that the 
accumulation of combustion products in the vicinity of the flame due to the absence of buoyant 
force (and hence, reduction in entrainment and mixing) makes diffusion the dominant mechanism 
of transport and results in longer residence time for fuel to burn. The normal-gravity heights are 
the averages of minimum and maximum flickering-flame height. In general, the lower the oxygen 
concentration, the taller the flame, regardless of gravity level. This is obvious since the residence 
time for burning the fuel is related to the level of oxygen in the environment. 

Figure 10 shows that at some high-enough oxygen concentration, the normal-gravity and 
microgravity flame lengths may become equal. This indicates that oxygen may be so abundant that 
accumulation of the combustion products in the vicinity of the microgravity-flame front may have 
no effect on the diffusion of oxygen toward the flame. This phenomena has not been observed for 
propane flames (see Figure 1 1), probably due to massive sooting, tip opening, and quenching at 
the tip even at high oxygen concentrations. 

Significant changes in flame color and luminosity occur in the absence of buoyancy for 
different oxygen levels. In microgravity, the oxygen concentration has a strong effect on methane 
flame heights, whereas color is more significantly changed in propane flames (see Figures 10 and 
1 1). In addition, reduced buoyancy makes the flames wider compared to normal-gravity flames. 
This is due to the lack of entrainment, continuous release of products, and dominance of diffusion, 
which are responsible for the adjustment of the flame to the location of stoichiometry. However, 
in normal gravity, the buoyant force aids the arrival of oxygen to the flame front causing the flame 
to be pencil-like. 

The various colors observed in microgravity have significant physical implications. In 
general, methane flames do not change drastically in color upon the removal of buoyancy, although 

18 




Fig. 9. Effects of pressure and oxygen concentration on microgravity laminar propane diffusion flames burning in 
quiescent oxygen -nitrogen environments. Here, the nozzle has a diameter of 0.15 cm, and the fuel flow rate 
corresponds to 1.0 cc/s under the standard conditions of 1 atm and 294 K; the cold jet Reynolds number 
based on the fuel properties and nozzle radius is 98. Silicon carbide fibers (15 ^m in diameter) were 
stretched across some of the flames to identify the flame location and approximate level of temperature. 
Note that the normal-gravity counterparts of these flames were almost identical in color (i.e., were yellow), 
had a closed tip, and were flickering. The extensive red observed in the flame of 30% Ch-0.5 atm is 
reflection from the chamber wall. Reproduction has increased the color intensity for yellow and red, and has 
reduced the intensity for blue. 



19 





Fig. 10. Normal-gravity and microgravity flames of methane at 1 atm as a function of oxygen concentration The 
nozzle had a tapered-tip radius of 0.048 cm. The various colon indicated in the diagram are as follows- 
B (blue). D (dark), O (orange), P (pink), R (red), W (white), and Y (yellow). The bars show the range of 
normal-gravity flame flicker (0; reproduced from [18]. 



20 



Proptnt Fl«m< 
181 Oxygen 
1 ICT. 

lirge Nozzlt 
0.96 cc/s 




1-9 



Propint Hunt 
ZlS Oxygen 
1 <tn. 

Urgt Nozzle 
0.96 cc/s 




1-9 



0-9 



Propint FIwm 
301 Oxygen 

1 109. 

Sntll Nozzlt 
0.96 cc/s 



.w/p/l 



-1± 




Scale: 



1 cm 



u 



u 
1 • 

\> 
1. 








u 



u 



20 n 



j* 



H 



Fig. 11. Normal-gravity and microgravity flames of propane at 1 atm as a function of oxygen concentration. The 
nozzles had tapered-tip radii of 0.048 and 0.074 cm, respectively. The low-oxygen-concentration flame of 
propane in normal gravity shows pockets of flame leaving the flickering part The various colors indicated 
in the diagram are as follows: B (blue), D (dark), O (orange), P (pink), R (red), W (white), Y (yellow), and 
DR (dark red). The bars show the range of normal-gravity flame flicker (0; reproduced from [18]. 



21 



they become dimmer due to the larger flame surface, resulting in more radiative loss, and hence, 
cooler flames. However, the propane flames in microgravity change remarkably in both color and 
characteristics. Referring to Figure 9, the dim, large blue flame of propane at 18% oxygen 
concentration shows a cooler overall flame temperature with little or no soot formation. In 
addition, the intense yellow/ orange/red/dull red colors observed at higher oxygen concentrations 
show that extensive soot formation, burn-off and quenching occur with flames appearing 
underventilated even in the 30% - O2 environment Recent tests with 50% oxygen have still 
shown massive sooting with open tips (apparent undcrventilation), indicating that soot quenching 
may not be avoided in microgravity by merely increasing the oxygen level. 

The unique underventilated behavior of microgravity propane flames shows that extensive 
soot formation due to pyrolysis, increased residence time, and radiation cooling is a characteristic 
of microgravity flames. Indeed, the films show that soot escapes through the tip of the flame. 
This tip-quenching is believed to be due to limited oxygen-fuel mixing, radiative loss from the hot 
combustion products and soot particles, and, possibly, unbumed fuel escaping through the flame 
tip. The low-oxygen-concentration flames of propane (Figures 9 and 11) show that near-limit 
extinction may have been approached, and that the limit and mechanism may be different in micro- 
gravity compared to normal gravity. It seems that at low-enough oxygen concentrations, we may 
observe a flame of the same size but so cool and dim indicating that heat release due to reaction can 
no longer compete with heat loss, causing flame extinction in microgravity, but not in normal 
gravity. 

The 18% - O2, 1-atm flames of propane in Figures 9 and 1 1 look different, i.e., Figure 9 
shows some red developing inside the flame. These tests were conducted in the 2.2-Second Drop 
Tower, and show a transient behavior in terms of a change in color but not shape and height. 
Recent 5.18-second microgravity tests of this flame have shown that during the first 2 seconds of 
microgravity, the flame is entirely blue. After 2 seconds, red (quenched soot emission) starts to 
develop inside the blue boundary, and increases in intensity throughout the rest of the drop period. 
This unusual behavior occurs slowly such that 5 seconds of microgravity is not sufficient to 
quantify the behavior of these low-oxygen-concentration and/or low-pressure flames. The 
increased residence time in microgravity flames affects the rate processes of soot formation, 
agglomeration, and burn-off, delaying the approach of the flame toward steady state. This subject 
will be discussed in more detail later. 

In order to study the effects of pressure on microgravity flames, tests were conducted at 
different oxygen concentrations for the range of 0.5-1.5 arm in pressure. Propane flames burning 



22 



in air are shown in Figure 12 as a function of pressure. Observations on the color of micrograviry 
flames show the strong effects of pressure on sooting, flame length, and flame diameter. 

Microgravity flame heights are plotted as a function of both pressure and fuel Reynolds 
number in Figure 13. It appears from these data that the flame height correlates linearly with 
Reynolds number, but reaches an apparent minimum between 0.5 and 1.5 atm. The flame height 
in all of these tests was based on the location where red soot changes color into dull red, indicating 
local flame extinction (as can be seen in Figure 12). Experimental studies have shown that soot 
ceases to burn near 1300 K. The red to dull-red transition seems to roughly correspond to this 
temperature (as was discussed in Section ITJ.a). It is believed that red/dull-red boundary gives a 
more accurate measure of flame height. 

Figure 14 shows the soot quench region (defined as the length of the red zone at the flame 
tip), which increases with both pressure and fuel Reynolds number, indicating enhanced soot 
formation. This requires a longer residence time for soot to quench. The lowest Reynolds number 
at 0.5 atm shows no quench length since the flame tip is closed. 

In Figure 15, the blue-base height (i.e., soot inception and agglomeration period prior to 
emission and bum-off) is plotted for both microgravity and normal-gravity flames as a function of 
fuel Reynolds number at different pressures. As the pressure increases, both normal- gravity and 
microgravity soot-inception heights decrease with little dependence on flow rate. 

Figure 16 shows the microgravity flame-tip diameter vs. Reynolds number. Due to the 
uncertainty in the beginning of tip opening at 0.5 atm, no attempt has been made to connect the 
three data points. The 1.5-atm and 1.0-atm flames show an asymptotic type of convergence, 
indicating that beyond a certain Reynolds number, the flame tip has a fixed diameter. Therefore, 
regardless of pressure (and hence, availability of oxygen), jet momentum and fuel-mass flow rate 
dominate the flame adjustment to the location of stoichiometry. 

These results show the unique characteristics of sooting microgravity flames. In general, 
for both normal-gravity and microgravity flames, the soot inception period and the regions of 
burning CO, C2 emission and CH emission (blue region at the base) decrease as the pressure 
increases. This indicates that formation, bum-off and emission of soot is facilitated by pressure 
increase. Normal-gravity studies of hydrocarbon flames have shown that a pressure increase 
results in an increase in soot formation and a decrease in soot removal by oxidation. This agrees 
with the present observations of the normal-gravity flames, since brightness increases due to more 



23 



1-g 
1.5 a tin 




v*i* 








*ttn 



Fig. 12. Normal-gravity and microgravity flames of propane in air with nozzle diameter ■ 0.15 cm and fi 
Reynolds number * 98 (based on the nozzle radius). The various colors observed are as follows: B (blu 
BB (bright blue), D (dark), DB (dark blue), DP (dark pink), DR (dark red), O (orange). P (pink), R (re- 
W (white), Y (yellow). The range of flicker (f) is also indicated for normal-gravity flames; scale: t— — 
1 cm, reproduced from [19]. 



24 



12 

10 

T 

£ 6 
* 

I . 

2 






4» 



M 

tt/neKi liaMr 



W7 



o.s 



1< 










10 

8 








it • 1*7^/ 


z » 

* 

i t 
I 








II^X*^ 


■ 








»• 4 








**^^ JC ^ 


2 















■* ' 


' 


• 



1.0 
Pr«siur« (tta) 



1.5 



Fig. 13. Microgravity propane/air flame heights vs. pressure and fuel Reynolds number. The lines connect the data 
points; reproduced from [19]. 




Fig. 14. Soot quench length (red region at the flame tip) vs. fuel Reynolds number as a function of pressure for 
microgravity propane-air flames. The lines connect the data points; reproduced from [19]. 



25 



1.2 
1.0 

I 0.1 

^ o.» H 

« 

S o.<|- 

' o.: - 



o.o 




-4—) 



0.5 tta 



1.0 iU 



l.S tta 



4* 



147 



Riyna 1 A» 



Fig. 15. Blue-base height vs. fuel Reynolds number at different pressures for normal-gravity flames (open symbols 
and dashed lines) and microgravity flames (closed symbols and solid lines). Whenever the normal-gravity 
and microgravity data coincide, only the microgravity symbol is used. The lines connect the data points; 
reproduced from [19]. 




la/mitt 



Fig. 16. Flame-tip diameter vs. fuel Reynolds number at different pressures for microgravity propane-air flames. 
The lines connect the data points. The broken line indicates uncertainty in connecting the data points due 
to the lack of information on Reynolds number at the onset of tip-opening; reproduced from [19]. 



26 



soot, and hence, more emission in normal-gravity flames. In microgravity, as the pressure 
increases, the length of the yellow region decreases, but orange, red and dull red regions increase 
in length (see Figure 12), indicating that more soot has been formed 

The complex processes of fuel pyrolysis and soot formation, agglomeration and bum-off 
are all pressure-dependent. A pressure increase results in an increase in collision frequency of the 
particles, and hence, growth due to agglomeration. Also, collision frequency depends on the 
particle size, which is a determining factor in particle bum-off. But particle bum-off is oxygen- 
dependent. Therefore, there exist different, oxygen-dependent, competitive rates of fuel pyrolysis 
and particle agglomeration and bum-off, which suggest the importance of the relative effects of 
these rate processes. The competition between these rate processes may be responsible for the 
non-monotonic behavior of the flame height with pressure in both normal gravity [19] and 
microgravity (see Figure 13). It seems that there exists a pressure in the range 0.5 atm < P < 1.5 
atm for any Reynolds number, at which the flame attains its minimum height due to the most 
efficient burning caused by the competing effects of oxygen availability, pressure, and the rate 
processes mentioned earlier. This phenomena is exacerbated by increasing Reynolds number. 
Hence, at low-enough Reynolds numbers, the flame heights may become equal regardless of 
pressure (see Figure 13). 

(d) Radiation Measurements 

Flame-radiation studies are being conducted in the 5.18-Second Zero-Gravity Facility for 
both methane and propane flames. In addition, normal-gravity counterparts of these flames are 
being studied simultaneously. Using a wide-view-angle, thermopile-detector radiometer, data is 
collected in terms of radiance (W/cm 2 -sr). Figure 17 shows a typical set of data (in mV) for flame 
radiation in both normal-gravity and microgravity environments. The results have shown that 
radiative loss (in terms of radiance) is generally 6-10 times larger in microgravity. This is due to 
the larger flame size and extensive sooting. In addition, examination of the data shows that 
radiation does not reach steady state in 5 seconds of microgravity. This is apparently due to the 
accumulation and slow transport of combustion products in the vicinity of the flame. The 
luminous region does not appear to increase in size (from visualization). However, the diffusion 
of non-luminous hot gases in the surrounding causes the volume of the hot gas to increase, which 
is reflected through the continuous increase in radiation. This has a direct impact on the approach 
toward steady state in microgravity flames, indicating mat reduced-gravity earthbound facilities do 
not provide enough time to obtain the true steady-state data. 



27 



12 



Thermal 
Radiation 
(ibV) 



8 




Drop 
Starts 



Tine (sec) 



Fig 17 Radiometer (thermopile detector) response from a propane flame burning in ^^"L 1 "" ^ 
h^h normal eravitv and (5 18 sec.) micrograviiy conditions. The average radiance obtained from these 
£u us™g r^^tweTn 4 sec. aS 5 ic.) is 0.65 W/cm*-sr for the nomud-g^jr flam^n 
5 32 W/cm 2 -sr forUie microgravity flame. Flame height = 5.0 cm (avg.) and 11.23 cm, and flame 
r^imum dimeter = 0.84 enfand 2.70 cm, for normal-gravity and micrograv.ty flames, respecuvely; 
reproduced from [20]. Fuel-flow rate = 1.5 cm 3 /s and nozzle radius = 0.0825 cm. 




VoIum Flow Ratt (ar/stc) 



Fig 18 Effect of fuel-flow rate on radiance from propane-air flames at 1 atm under both normal-gravity and 
microgravity conditions. The radiance data are the average values between 4.0 sec and 5.0 sec alter 
ignition; reproduced from [20]. 



28 



The data presented here is the first demonstration of flame-radiation measurement in 
microgravity environments. One important finding is that radiative loss may become a very crucial 
factor in radiative ignition in low-gravity environments, namely, a small fire may spread much 
more rapidly in micrograviry because of the significantly higher levels of radiation. The data of 
Figure 18 support this argument. In normal gravity, a large fraction of heat release is removed 
rapidly via the convection of hot products of combustion, and therefore, does not contribute to the 
process of radiative loss from the flame zone. In microgravity, practically all of the released heat is 
accumulated around the flame through the combustion products, which causes radiative loss of up 
to an order of magnitude higher compared to normal gravity condition. 

Pressure and oxygen concentration also show a very significant effect on radiative loss in 
microgravity. Figure 19 shows radiation from methane flames in both normal-gravity and micro- 
gravity environments under different conditions of pressure and oxygen content. The normal- 
gravity data show no sensitivity to the environmental condition, whereas radiation levels from 
microgravity flames are strongly affected by pressure and oxygen concentration. Figures 20 and 
21 show these effects in more detail. 

One interesting finding from the radiation data is that low-pressure/low-oxygen-concentra- 
tion flames show near-steady-state radiation levels by the end of the available 5 seconds of micro- 
gravity, whereas high-pressure/high-oxygen-concentration flames show a continuous increase in 
radiative loss all the way to the end of the drop, indicating that 5 seconds of microgravity is not 
sufficient to obtain steady-state data in these flames. Figures 22 and 23 show the effects. 

Figures 22 and 23 also show that the low-pressure/low-oxygen-concentration flames that 
are entirely blue (and therefore essentially soot-free), show appreciable amounts of radiation which 
would normally be associated with CO2 and H2O bands. Therefore, the radiation contribution 
from gas-phase combustion products is as important as that of soot in microgravity flames. 

Figure 24 shows that even at 15% oxygen, a very faint, blue microgravity flame of 
methane still radiates appreciably. However, it is not obvious what longer test times would show 
for these very weak flames. 

(e) Temperature Measurements 

Quantitative data including the temperature field (using a fixed rake of nine thermocouples) 
have also been obtained in the 5.18-Second Zero-Gravity Facility for methane and propane 



29 



1 atm 



21% 0- 



12 



E 



C 
o 



A3 




t (sec) 



12 



P = 1/2 atm 
21X 2 



P = 1 atm 
185 0, 













































1-g 


o-j 










-r-fA^— 


ty-'-i/Vat 


^^ 






. r 


1 


J_>>= 



t (sec) 



-12 


I 










E 















C 

o 










0-9 




4-> 

10 






~/\_*/- 








NS 








i-g 














t » 













6 



t (sec) 



Fig. 19. Effects of pressure and oxygen concentration on radiative loss from methane flames under both normal- 
gravity and microgravity conditions. Here, fuel-flow rate * 3.0 cm 3 /s, and nozzle radius * 0.0825 cm. 



30 



o.u 










/ • 


5.0 






o-g 


• / 




4.0 












3.0 












2.0 












1.0 


- 






/'■' 


4 








• 








1 




1 i ,,.. 1 - 


• 



0.0 



15X 



21X 
Percent Oxygen 



30X 



Fig. 20. Effects of oxygen concentration on radiative loss from propane flames. Here, P = 1 atm, fuel volume 
flow rate = 1.5 cm 3 /sec, and nozzle radius * 0.0825 cm. 



5.0 



4.0 _ 



~ 3.0 

U 

c 
I 

CM 

4, 2.0 I— 

g 

1 

1-0 U 



o.o 



i-g 



J. 



30X-0, 




1BX-0, 



i-* 



30t-Oj ; l8X-0 2 



_L 



J_ 



0.5 1.0 1.5 

Pressure (ata) 



Fig. 21. Effects of oxygen concentration and pressure on radiative loss from methane flames at a fixed fuel mass 
flow rate corresponding to 3.0 cm 3 /sec at 1 atm; nozzk radius « 0.0825 cm. 



31 




(b) 




Ttmt (*«) 



Fig. 22. Radiation from normal-gravity and microgravity flames of methane burning in (a) 0.5-atm. 18%-0 2 , and 
(b) 1.5 atm, 30%-O2 environments. The fuel mass-flow rate is fixed corresponding to 3.0 cm 3 /sec under 
atmospheric condition; nozzle radius = 0.0825 cm. 




J 4 

Time (uc) 



(•«) 



Fig. 23. Radiation from normal-gravity and microgravity propane flames in (a) 0.5-atm, 18%-C>2, and (b) 1.5 atm, 
30%-O2 environments. The fuel mass-flow rate is fixed corresponding to 1.5 em'/sec under atmospheric 
condition; nozzle radius * 0.0825 cm. 



32 





12 



10 



O 
CE 



i l i I 





Fig 24. Flame radiation for both normal -gravity and micrograviiy flames of methane (shown in the photographs) at 
0.5 atm and 15% oxygen; nozzle radius = 0.0825 cm; mass-flow rale corresponds to 3.0cm 3 /scc under 
atmospheric condition. Reproduction has changed yellow to white, and has reduced the intensity and 
length of the blue flame. 



33 



diffusion flames burning in quiescent oxidizing environments of various pressure and oxygen 
concentrations. 

Figure 25 shows the size, shape, and color of some of the flames studied in addition to the 
position of the thermocouple rake with respect to the flame. The micrograviry flames shown in this 
figure correspond to the shapes and colors observed just prior to the deceleration of the package. 

Figure 26 shows a typical set of temperature data for the propane flame burning in 19%-C>2/ 
81%-N 2 environment at 1 atm, under both normal-gravity and microgravity conditions. The 
lowest row of thermocouples is 10 cm above the nozzle tip. An initial overshoot in temperature is 
observed in both cases due to the presence of excess fuel at ignition. This figure presents some 
useful information not only in relation to the temperature field, but also for other flame 
characteristics. The normal-gravity data show that the temperature drops quickly to the ambient 
somewhere between the radial locations of 0.75 and 2.75 cm. This is due to the effect of buoyant 
force which carries and removes the combustion products in a narrow column determined by the 
maximum flame radius. In addition, the strong buoyancy-driven portion of the flame causes the 
three thermocouples near the centerline to show roughly equal temperatures, which essentially 
reach steady state approximately 3 seconds after the start of the experiment. 

The data presented in Figure 26 for microgravity flames show that far above and away 
from the flame, the gas is still experiencing a temperature rise due to the continuous release of the 
combustion products in the vicinity of the flame and their subsequent diffusion. This figure also 
shows that the gas temperature does not reach a steady value anywhere in the field during the 5 
seconds of microgravity, again due to the continuous dilution and heating of the environment 
caused by the combustion products. In addition, large temperature gradients exist in the axial 
direction for the microgravity flame, and the variation of temperature in the radial direction is 
significant all the way to the far field. 

Figure 27 shows the temperature data at t s 5 seconds as a function of the axial location of 
the thermocouples for fixed radial locations. Both normal-gravity and microgravity data are shown 
in this figure. All of the normal-gravity data indicate that the second and third columns of 
thermocouples (see Figure 25) register a temperature of 300 K. The first column (closest to the 
centerline) shows a constant and small decrease in temperature with an increase in height for more 
diffuse flame due to the lower pressure. The microgravity flames A and B of Figure 27 show 
almost identical temperature distributions. Significant differences in temperature levels are 



34 




1-10 pn 




P - 0.5 aim 
21VO, 
L-i em 




l-»— . 




P ■ i aim 

21VOj 

L-iOcm 




U-4- 



I ; 



!-♦ 1- 



"Tl 






*¥£*-. 



£1 

p ■ i •» 

into, 

L-lcm 



rt 1 am 



Fig. 25. Flames of propane at 1.5 cm 3 /sec for both normal-gravity and microgravity conditions. The colors are as 
follows: B (blue), D (dark), DB (dark blue), DR (dull red), O (orange), P (pink), R (red), W (white), 
Y (yellow). The flicker range, f, is shown for the normal-gravity flames. L is the distance between the 
nozzle tip and the lowest row of thermocouples. The columns of thermocouples are, respectively, 
0.75 cm away from the centerline, and then 2 cm apart The rows of thermocouples are 3 cm apart 
These diagrams are to scale; reproduced from [21]. 



35 



0.7S 



\i 



HOT TO 
SCALE 



.Iain 






t-- 



l • 10 



Propant -195 2 Flame 
Flow Rate -1.5 cc/s 
Nozzlt Radius • 0.0825 ca 

P • 1 etw 

See Fig. 25(B). 



Normal Gravity 



Hicrogravity 



12CC 



20: 



^SE 



t (sec) 



-LA. 



*^ 



uoo 



200 



m 



^ 



t (stc) 



2" 



_P : __ == __J ~-^ 



t (sec) 



1200 



200 



k 


H 




_ r «- 


r^p-^ 



t (stc) 



12:: 



200 

















































































Q 












~TT~ 



















t (stc) 



use 



200 

















































































h 


a 














■■•— 










1 





t (stc) 



Fig. 26. Typical temperature data for a normal-gravity and microgravity Dame; reproduced from [21]. 



36 



observed between microgravity flames C and D, with the 0.5-atm flame being taller and wider. 
The combination of larger flame size and reduced pressure is responsible for this behavior. 

Figure 28 shows the temperature data as a function of the radial location of the thermo- 
couples for fixed axial locations in the microgravity flames. The normal-gravity data is not 
presented in this figure, since the temperature drops to ambient somewhere between the first two 
columns of thermocouples. Due to the uncertainty in the monotonic behavior of temperature 
between the first and second thermocouples of the lowest row, the plots have not been extended to 
the temperature at the first thermocouple. It is possible that there is a peak in temperature between 
the innermost thermocouples due to the presence of the flame zone, especially for flames C and D. 

The temperature data, which like radiation-measurement results, are the first demonstration 
of the ability to map the far field in a microgravity combustion study, do not only provide 
quantitative data for comparison with the predictions, but also are a means of gaining insight into 
the characteristics of flames in addition to their behavior in microgravity environments. The data 
presented here show that 5 seconds of microgravity are not sufficient to obtain information on the 
flame structure, since the thermal field is continuously changing, at least in the available time of 
five seconds. In addition, the data suffers from the lack of capability to obtain a map of the near- 
field temperatures including the flame front. Longer test times provided by a space experiment 
would allow the complete mapping of the temperature field. 

(f) Predictions 

The mathematical model of [10] is under modification to obtain solutions for laminar jet 
diffusion flames. The approach involves a finite-difference solution of the partial differential equa- 
tions for mass, momentum, elements, and energy, which are in their boundary- layer form. The 
effects included are diffusion, viscosity, inertia, kinetics, gravity, and radiation. The current 
chemistry assumed is that of shifting equilibrium. 

In this axisymmetric, steady-state, vertical laminar-jet configuration, the fuel is injected at a 
finite velocity into the quiescent surroundings. The effects upon the velocity field are shear or 
molecular diffusion of momentum, pressure gradient, and gravity. The buoyant force is a 
combination of gravity and pressure gradient. In the vertical-jet problem, convection of energy, 
heat release rate, heat conduction, and the energy transported due to diffusion of species affect the 
temperature field through the velocity field and the concentration field. The concentration field is, 
in rum, affected by the convection of species through the flow and therefore by the velocity field. 



37 



1200 
1100 

1000 

900 
800 
700 
600 
500 
400 
300 



P ■ 0.5 atm 
211 2 
L • 8 cm 




P • 1 atm 
211 02 
L • 8 cm 




11 



14 



J- 



11 14 10 

Axial Location (an) 



P • 1 *tn 
191 02 
I • 10 OK 




P ■ 1 itfll 

211 0j 
L ■ 10 cm 




13 



16 



10 



U 



16 



Fig. 27. Temperature plots as a function of the axial location of thermocouples for both normal-gravity (open 
symbols and dashed lines) and microgravity (closed symbols and solid lines) flames of propane with a 
volume flow rate of 1.5 cm 3 /sec. L is the distance between the nozzle tip and the lowest row of thermo- 
couples. The radial locations indicated in the diagram are applicable to all plots; reproduced from [21]. 



1200 


■ D 


P » 0.5 ata 
211 2 


• 


c 


P • 1 atm 
211 O2 


- B " l ' ta 

191 O2 


. A 


P ■ 1 tUH 

211 2 


1100 




L « 8 cm 


■ 




L • 8 an 


L • 10 CM 




L • 10 an 


1000 


• 
















900 












- 


» 




aoo 




Axial Location 




• 










700 




v X « 8 » 








Axial Location 


• 




600 


" X« 11 cm^ 






\ 


\ 


X • 10 c» 


v 


\ 


500 








\ 


s\ 


J«13 OlNyX 


N 


\ 


400 


X-14 cm 






V 


^ 


j'lToNX 


'V 


^ 


300 










^•^Ofc 


^^ 




^Hfc 


i 


' 


i 


1 




- 


' 


' 






-rf— i 




-*- 








4 ■ 





0.75 2.75 4.75 0.75 2.75 4.75 0.75 2.75 4.75 

Radial Location (ca) 



0.75 2.75 4.75 



Fig. 28. Temperature plots as a function of the radial location of thermocouples for microgravity flames of propane 
with a volume flow rate of 1.5 cm 3 /sec. L is the distance between the nozzle tip and the lowest row of 
thermocouples. The axial locations indicated in the diagram are applicable to those of the same family 
(i.e., same L). The incomplete plots indicate the presence of flame between the two columns of 
thermocouples; reproduced from [21]. 



38 



Chemical production rate and diffusion are the mechanisms which directly cause changes in the 
species field. The diffusional gradients produce an air/fuel ratio distribution ranging from fuel rich 
to fuel lean in the mixing region. The surface upon which the equivalence ratio is unity is referred 
to as the "flame front", which corresponds to the maximum temperature. The assumptions made 
are: (a) diffusive processes are important only in the radial direction, (b) combustion is diffusion- 
controlled, and (c) flow is steady and axisymmetric. The equations of continuity, momentum, 
diffusion of elements, and energy are nondimensionalized with respect to the state of the flow at 
the maximum flame width point. The equations of state are used to complete the analysis, and 
radiation has recently been incorporated to include water and carbon-dioxide contributions via the 
approximation of optically thin medium. 

The governing flow-field conservation equations are transformed into the von Mises plane, 
which eliminates the global continuity equation, and then solved numerically by employing an 
explicit finite-difference technique. The details of the flow field, i.e., velocity, temperature, and 
species fields along with the gross characteristics including combustion lengths and flow 
deflections are obtained by the solution of the system of conservation equations. 

The model is currently being modified to incorporate the effects of: (a) soot formation and 
burn-off, (b) radiation from soot in addition to gas-phase radiation by incorporating the radiative- 
transfer equation in the formulation, (c) thcrmophoretic effects, which become significant in 
microgravity, (d) quasi-global kinetics, (e) multi-component diffusion in the presence of soot, and 
(f) elliptic effects due to the significance of axial diffusion in microgravity. 

Experimental results have revealed the non-uniform heat-release mechanism throughout the 
flow field; partial combustion and soot formation dominate in the near-jet region, while the soot 
bum-off process is facilitated downstream due to increased residence time and continued heating. 
The kinetics of the combustion process under microgravity conditions significantly influence the 
accuracy of the predictions as can be seen in Figure 29, which shows the effect of heat release on 
flame shape and its comparison with the experimental data. The partial oxidation analysis shows 
smaller flame radii and shorter flame lengths. Further results are compared with the experimental 
data in Figure 30; they show substantially improved agreement with respect to the results obtained 
assuming complete combustion model. This confirms the potential role of chemical kinetics in 
microgravity flames. 

Figures 31-33 show other examples of application of the model, where predicted micro- 
gravity flame heights are compared with experimental data [11]. A satisfactory trend in the 



39 





i 






,,. * 


■ Mrimint 




E 


1.6 
1.2 


\ 


* 
i 




Froi«n flow, CH 4 /oir only 
^. Full chemistry modal 






8 


- lY 


^ *» j 


- HjO ond CO only 








/ A 


i^, 


. HjO ond C(i) only 






4 














J^ l 


1 1 1 1 


| 




( 


)4 8 


12 16 20 2.4 


2 8 












/;cm 





Ooto 
- 1 .08 cm*/i 
R« z 40.9 
r = 294 K 
P s 1 otm 

r Q s 0.051 em 



Fig. 29. Effect of chemistry model upon analytical prediction of a steady-state zero-gravity methane-air flame- 
reproduced from [10]. The experimental data were obtained from [15]. 



20 



16 



12 - 

e 

4 



Com pn r« combustion model, 
M 2 0, C0|, CO formed-. 




10 



X» 



R«o 



Portiol Oiidotion 
theory 


r ,cm 


H ? 0, CO HjO, C(»l 




o N> 


0.051 


A 


0.0825 



1000 



Solid symbols d«nof« 
• ip«rim«ntal 00 m 



Fig. 30. Comparison of theoretical model with experimental zero-gravity flame widths in ambient methane-air 
flames, as a function of fuel Reynolds number; reproduced from [10]. The experimental data were obtained 
from [15]. 



40 

















12 




£*9- 


s.s. 


Futl 


(e.) 






10 


A 

m 






O.OSl 
0.062S 


A 



















i 



j_ 



_L 



X 



30 W 10 120 ISO MO 210 240 

layittHi MMtar 



Fig. 31. Predicted and measured microgravity propane-air flame heights at 1 atm vs. jet Reynolds number; 
reproduced from [11]. 





m 
















12 






S.S. 




tt 














^f* 


ThMr£ 


FmI 


(«) 




▲ 


^_ 


«4 


O.OSl 


10 


m 










■ 


— 


OU 


o.ous 












• 


• 














1. 


m 






/ 
/ 

/ 








5 

r 






4 
/ 
/ 


■ 


'* 






- 4 






t 










1 




/ 


'■ 


/* 








Z 
2 




/A 

















i 


1 




1 


■ i 


1 



JO W M 120 

ft*jfl»14» 



ISO 110 210 240 



Fig. 32. Predicted and measured microgTavity methane-air flame heights at 1 atm vs. jet Reynolds number; 
reproduced from [1 1]. 



41 



too 



i 100 



1 
I 




•»! 


ka 

:i5,i6 


■ g 

(11) 


an. 


A 


A 


la*'~ 


e 




»-»T'«- 


a 




mi 




♦ 



I }*•!■ 



(12) 
(11) 



-L. 



•oo no too *oo 

ttynoldt luabir 



Fig. 33. Measured and predicted microgravity flame heights for different fuels burning in air at 1 atm; reproduced 
from [11]. 



4 > 



:-» 



ll / 




» 1 / 


Vi 1^^^" 


' 1 / 


yr 




* 








4 

10-*. 




irH /»«*V/*« 



I («■) 



Fig. 34. Predicted non-dimensional centerline velocity (with respect to jet exit velocity) vs. axial distance along the 
jet as a function of gravitational level; methane-air flames, nozzle radius » 0.0825 cm, fuel-flow rate = 
1.0 cc/sec, pressure * 1.0 atm, and jet exit velocity ■ 46.8 cm/sec; reproduced from [1]. 



42 



behavior of the flames is demonstrated. However, the propane flames (Figure 31) are slightly 
longer and the methane flames (Figure 32) are shorter than the predicted flames. One explanation 
is the absence of soot and associated radiation in the model in contrast to the high concentrations of 
soot observed in most of the micTOgravity flames. In addition, the absence of elliptic effects (axial 
diffusion) and the relatively simple predictions for the molecular transport rates (i.e., species, 
momentum, and energy) may contribute to the differences between the predicted and measured 
flame heights. 

Figure 33 shows a comparison between different theoretical and experimental results. The 
data points correspond to the steady-state flame heights of [11], [15], and [16]. The predictions 
are those presented in Figures 31 and 32 as well as the results of another theoretical modeling 
effort [12]. The two numerical models show a good agreement with each other and with the data. 
The propylene flame heights of the previous studies [15] do not agree with the predictions of [12]. 
This is surprising since propylene and propane are very close in chemical structure, and 
satisfactory agreement has been obtained for methane, ethylene, and propane. However, the 
present study does not deal with highly sooting fuels such as propylene and ethylene. 

Application of this model to a family of methane flames under different gravitational levels 
shows the important role of convective effects (Figure 34). It can be seen that centerline velocity 
does not play a major role in diffusion flames under gravitational accelerations of <l(r 3 g. 

(g) Summary of Observations and Conclusions 

• Flames in microgravity are taller, wider, sootier, and more diffuse than their normal- 
gravity counterparts. 

• Low-pressure and/or low-oxygen-concentration flames in microgravity are entirely blue, 
unlike those in normal gravity. 

• High-pressure and/or high-oxygen-concentration flames show massive sooting in 
microgravity. 

. Flame-tip opening (underventilated-like behavior) has been observed in all of the 
microgravity propane flames studied to date. 

• Most of the methane flames do not have open tips in microgravity. 

• Radiative loss for all of the normal-gravity flames studied falls in the range of 
0.0-1.0 W/cm 2 -sr. 



A3 



• Microgravity flames show a radiance of 6- 10 times larger than those in normal gravity. 

• Faint, blue flames of low-oxygen/low-prcssure environments show appreciable amount 
of radiative loss compared to their normal-gravity counterparts. 

• Pressure and oxygen concentration have a significant effect on radiation from 
microgravity flames, but not from normal-gravity flames. 

• Radiation does reach near-steady state for low-oxygen/low-pressure flames in 
microgravity, however, other flames show a continuous rise in radiative loss during the 
5 seconds of microgravity. 

• Continuous change in color with time is observed for atmospheric-pressure, low- 
oxygen-concentration flames in microgravity; the color is blue at the beginning of the test 
followed by a steady buildup of red inside the flame up to the 5 seconds of the available 
test time. 

• Appreciable temperature rise has been measured in the far-field of microgravity flames. 

• Much steeper temperature gradients exist along the flame centerline in microgravity due 
to the lack of buoyancy. 

• Temperature data do not show an approach toward steady state in 5 seconds of 
microgravity. 

Of these observations, the most significant is the effects that pressure, oxygen 
concentration, and fuel type have on microgravity flames in contrast to the effects of these para- 
meters on their normal-gravity counterparts, and that thermophoretic effects, diffusive processes, 
radiative loss, soot formation and oxidation, and gas-phase kinetics are critical to the description of 
these microgravity flames. 

Based on this conclusion, the specific objectives for the continued work have been defined. 

IV. OBJECTIVES 

The overall objective of this program is to study the effects of pressure, oxygen 
concentration, and fuel type on the characteristics and structure of laminar gas-jet diffusion flames 
in the absence of buoyancy. 

As described in detail in the previous section, the data from recent ground-based 
experiments have enhanced our understanding of laminar diffusion flames in microgravity 



AA 



environments. However, unexpected characteristics have also been observed, raising new 
questions on transport : xresses in diffusion flames in the absence of buoyancy. 

The questions that need to be answered are: 

(a) What are the steady-state characteristics of the micrograviry flame, and given enough 
time, do the weak flames extinguish by virtue of their size, radiative loss, and 
associated transport processes? 

(b) Why does the removal of buoyancy result in underventilated-like behavior in some 
otherwise overventilated diffusion flames? 

(c) Why does microgravity flame length vary nonmonotonically with pressure, and what 
are the roles of buoyancy and oxygen concentration in defining this behavior? 

(d) Why are sooting characteristics sensitive to oxygen concentration in microgravity 
flames? 

V. HYPOTHESES AND DATA REQUmEMENTS 

This section discusses hypotheses for each of the questions raised in the prior objectives 
section, and identifies the data required to validate the hypotheses. 

(a) What are the steady-state characteristics of the microgravity flame, and given enough 
time, do the weak flames extinguish by virtue of their size, radiative loss, and 
associated transport processes? 

Hypothesis : The continuous release and slow transport of the combustion products 
creates a transient field in the vicinity of the flame during the early stages of burning 
(as is evident from the 5-second microgravity tests). However, a quasi-equilibrium 
state can be reached for both temperature and species fields, such that the flame 
approaches steady state, and the far field will not have an effect on the near-field (of 
the flame). This quasi-steady-state will persist for a period of time, and will then be 
followed by a transient state, in which oxygen arrival to the flame becomes more 
difficult, and the flame starts its transient approach toward extinction. On the other 
hand, for the weak microgravity flames of low-oxygen-concentration/low-pressure 
environments, there are two possibilities. The first is that radiative loss cannot 
compete with the heat release and consequent increase in temperature of the near-field 
region of the flame, again, resulting from the slow transport of the combustion 
products. In this case, given enough time, the flame may behave like the 18%-0 2 , 



45 



1-atm flame (see Section III) resulting in the formation and subsequent glowing of 
soot, and leading to a bright, sooty flame. However, the second possibility is that the 
heat loss by radiation exceeds the heat input to the near-field, and the flame becomes 
incapable of sustaining combustion, undergoing a rather slow extinction process. 

(b) Why does the removal of buoyancy result in underventilated-like behavior in some 
otherwise overventilated diffusion flames? 

Hypothesis: The removal of buoyancy increases the residence time available for soot 
formation within the flame zone. This increased residence time augments the pyrolysis 
of the fuel, leading to increased soot nuclei and growth. Larger soot and more soot 
formation result in slower oxidation and persistence of the soot. All of the soot is 
formed within the flame, and these larger particles (unlike gas-phase species) virtually 
do not diffuse. But these particles being inside the flame, are forced to move toward 
the flame centerline through the process of thermophoresis which is caused by 
temperature gradient. As a result, the particles core up and stay within a column 
defined by the width of the flame at its open tip, impede oxygen transport to the center 
of the flame, and prevent the flame from closing on its axis. It is interesting to note 
that microgravity methane flames do not show this underventilated-like behavior, 
whereas all of the propane flames studied to date (except very small flames, ~l-2 cm 
in height) have open tips. 

(c) Why does microgravity flame length vary nonmonotonically with pressure, and what 
are the roles of buoyancy and oxygen concentration in defining this behavior? 

Hypothesis: The observed increase in length with pressure is due to increasing 
amounts of glowing and burning soot rather than a change in the gas-phase flame 
height. At low pressures, kinetics effects tend to lengthen the apparent gas-phase 
flame zone. As the pressure increases, reaction rate increases and flame length 
decreases. As the pressure is increased further, soot levels increase, and this heated, 
glowing and burning soot persists beyond the gas-phase flame front. This gives the 
indication (through direct photography) of an apparent longer flame. Although the 
normal-gravity counterparts of these flames show the same non-monotonic behavior 
for (the average between minimum and maximum flickering) flame height, no 
conclusive information can be obtained from these normal-gravity flames. The reason 
is the high degree of flicker, in addition to the (generally) continuous decrease in 



46 



maximum flame height with increase in pressure and the non-monotonic behavior of 
the minimum flame height as a function of pressure [19]. However, the microgravity 
flames are flicker-free, and provide the opportunity to substantiate the above 
hypothesis in relation to this non-monotonic flame-height behavior. 

(d) Why are sooting characteristics sensitive to oxygen concentration in microgravity 
flames? 

Hypothesis : The microgravity environment provides longer residence times, resulting 
in prolific sooting. Higher concentrations of oxygen facilitate the burning of soot, 
causing higher temperatures. On the other hand, lower oxygen concentrations produce 
relatively cooler flames, reduced pyrolysis and soot formation, and reduced heat and 
mass transfer. The competing effects of increased residence time and relative transport 
rates of mass and heat due to Schmidt-number and Lewis-number effects will enhance 
the influence of oxygen concentration on these processes. 

To substantiate the hypotheses presented in this section, certain measurements are 
critical. These measurements are referred to as "required" measurements in the follow- 
ing discussion. In addition, there are measurements which greatly enhance the science 
return of this experiment, and each one helps in further validation of one or more of 
the hypotheses. These measurements are referred to as "desired" measurements. 

The following measurements are "required" for this experiment: 

(i) Temperature Field - Mapping of the temperature field is required for hypotheses 
(a), (b), (c), and (d). The temperature field must be defined because it is essential to 
fuel pyrolysis and thermophoretic transport of the soot particles. In addition, regions 
of soot quenching, exact flame location, far-field temperature map, and quantitative 
approach to steady-state flame characteristics will be obtained through temperature 
measurements. Also, understanding of the degree of completion of combustion and 
quantification of the radiative transfer of energy out of the flame will be accomplished 
by mapping the temperature field. When combined with the "desired" measurements 
for species field and soot size and concentration, the temperature field provides a map 
of effective emissivities and absorptivities which are essential to the modeling of 
radiative transfer. The mapping of temperature field also provides a means of 
validating the theoretical model. 



47 



(in Flame Visualization rCinematop-aphv^ - Visualization via cinematography is 
essential to monitor the behavior, color, luminosity, tip characteristics, development, 
and approach to steady state for all of the flames, regardless of the nature of the 
hypothesis presented in this section. 

iiiD Radiation Measurement - Flame radiation is an integral part of the 

measurements. It provides quantitative information on the flame development and 
approach toward steady state in addition to the variation of the thermal field with time. 
Through simultaneous measurements of overall radiation, and CO2, H2O, and soot 
radiation, a complete breakdown of contribution of gas-phase combustion products 
and particulates to radiative transfer can be obtained. This measurement is required for 
all of the hypotheses (a)-(d). In addition, radiation data help in developing the 
radiation submodel and validating the numerical model for flame characteristics. 

(iv^ Pressure Measurement - Measurement of pressure is needed to monitor any 
pressure rise that may occur due to the burning process, and to obtain estimates of the 
overall heat-release rate. This measurement is required for all of the hypotheses 
(a)-(d). 

The following measurements are "desired" since they enhance the data and help in 
further substantiation of the hypotheses. In addition, they provide unique data in 
characterizing the structure of microgravity diffusion flames. 

(v) Acceleration Data - Measurements of acceleration are desired to obtain the 
flame response to any residual acceleration or g-jitter effects during the experiment. 
These data will be used as input to the theoretical model for the prediction of flame 
characteristics. The acceleration data are equally important for hypotheses (a)-(d). In 
addition, theoretical analyses have shown that acceleration levels less than lO 3 g do 
not significantly affect the flame behavior. To validate these findings, acceleration 
measurements are desired. 

(vi'l Species Field - Measurements of the gas-phase species are desired to provide 
distributions of major combustion products, oxidizer, and any unbumed hydrocarbon. 
This is important especially for hypotheses (b) and (c), since it not only provides 
information on the distribution of combustion products, but helps to delineate the 
characteristics of the underventilated-type behavior (hypothesis b). In addition, the 



48 



influence of oxygen concentration on the burning process in microgravity will be 
quantified through the measurement of distributions of oxidizer and gas-phase 
combustion products (hypothesis d). 

(vii) Infrared Imaging for Species - This measurement is desired because a 
complete map of C0 2 and/or H 2 throughout the test period will be obtained. It not 
only identifies the flame location (locus of maximum products concentration) for 
hypotheses (a) to (d), but quantifies the development and transport of the field of 
products, thereby helping in validation of hypothesis (a). The issue of masking of the 
flame tip by soot will be addressed. In addition, for blue flames, the question of open 
or closed tip (given enough time for the flame to develop) will be answered 

fviiil Density-Gradient Field Imaging - This measurement technique is desired 
since a map of density gradient (and therefore, location of the flame) and qualitative 
field of temperature will be obtained. In addition, the flame response to disturbances 
caused by intrusive probes will be identified It also provides a map of any large-scale 
structures which cannot be obtained through direct visualization. This method 
provides information for hypotheses (a) to (d). 

fix') Velocity Field - This measurement is desired since it provides a map of the 
gas-phase velocity, helps to quantify the convective effects due to the initial jet 
momentum, and defines the nature of any stagnation region at the flame tip that may be 
contributing to soot accumulation and underventilated-type behavior. It is most 
important for hypotheses (a), (b), and (d). Model validation will benefit from 
velocity-field measurement. 

(x) Soot Size - This desired measurement provides the soot size at various 
locations in the flame. It helps in validation of hypotheses (a), (c), and (d). It 
provides information on soot processes (formation, agglomeration, bum-off) and 
thermophoTetic effects, in addition to soot build-up at the flame tip. 

(xi) Soot Concentration - This method is desired to obtain the field of soot 
concentration, to quantify the thermophoretic process, and to obtain information on the 
build-up and coring effect of soot (as described before). It helps to validate 
hypotheses (a) to (d). The data will also be used to improve the soot-processes and 
radiation submodels in the numerical model. 

In the next section, justification for a space experiment is presented 

49 



VI. JUSTIFICATION FOR LONGER-DURATION EXPERIMENTS 

Observations, measurements, and theoretical analyses show that five seconds is not 
sufficient to quantify the steady-state characteristics of microgravity laminar diffusion flames. 

Specifically: 

(1) Observations on the flame behavior show that although flame shape appears to reach 
near-steady-state, flame color continuously changes throughout the five seconds of 
available time. This indicates a continuous change in transport and chemical 
processes. 

(2) Temperature data show a continuous increase in the five seconds of microgravity, and 
that even the far field is being affected by the temperature rise, indicating a continuous 
release and show transport of the combustion products in the vicinity of the flame. 

(3) Radiation data show that although very weak (i.e., blue) flames show an approach 
toward steady state for radiative transfer in the five seconds of microgravity, brighter 
and sootier flames show a continuous increase in radiation loss, needing more time to 
reach steady state. 

(4) The results of a transient, analytical model for laminar diffusion flames in microgravity 
show that the flames need more than five seconds to reach near-steady-state. This is in 
direct agreement with the radiation and temperature data, and even with observations 
through visualization. 

In addition to the need for longer test times based on the results obtained to date from the 
5-second tests and analyses, longer-duration experiments are needed for each objective of the 
program presented in Section IV, as shown below. 

(i) In relation to Objective (a), i.e., the question of steady-state characteristics and the 
behavior of weak flames in microgravity, five seconds is not sufficient to allow the 
bright flames to approach steady-state, or the weak flames to show either steady-state 
approach or extinction process. 

(ii) In relation to Objective (b), i.e., the underventilated behavior of microgravity flames, 
it is possible that soot formation and subsequent quenching and accumulation at the 



50 



rip of the flame will ultimately result in extinction. This process may happen if soot 
stays at the tip, does not diffuse away from the flame, and results in reducing the 
flame height. In this case, the flame may get shorter in time until abundance of soot 
causes flame extinction. This requires longer microgravity test times. In addition, it 
is possible that some of the flames which had a closed tip in the 5-second tests may 
open given longer test times. 

(iii) In relation to Objective (c), i.e., the nonmonotonic behavior of micro-g flame height 
with pressure, it is conceivable that at longer times, this behavior may correct itself 
due to the interplay between the soot formation and accumulation, and the build-up 
and transport of gas-phase products. Longer times may result in abundance of soot at 
the tip of the flame, changing the nonmonotonic behavior. 

(iv) In relation to Objective (d), i.e., the sensitivity of sooting characteristics to oxygen 
concentration in microgravity flames, as the products of combustion continuously 
diffuse toward the far field, the effect of oxygen partial pressure may become less 
significant in the near-field, resulting in a different behavior of the flame compared to 
what has been observed in five seconds of microgravity. 

In the course of the current program, observations from the 2.2-second tests have led to the 
conclusion that 5-second microgravity tests should be conducted. However, measurements and 
observations from the 5-second tests have shown that still longer test times are needed. The results 
of theoretical modeling (see Figure 34), a separate analysis for the g-jitter effects on laminar 
flames, and recent tests of both bolted-down (~10" 2 g) and free-float (~10" 3 g) configurations of 
the frame for this experiment in the KC-135 aircraft, all point out to the fact that g-jitter and g levels 
in the aircraft will probably affect the flame, and that available times are not sufficient to 
characterize the flame. In the immediate future, we are planning to conduct tests in the KC-135, 
but based on the g-jitter analyses, the results are uncertain. The true understanding of the behavior 
and structure of these flames needs a combined set of advanced diagnostics for an extended period 
of time (of the order of 1-2 minutes). Regardless of whether these diagnostics can be used in the 
available earthbound facilities, the time constraint does not allow achieving the objectives of this 
program and substantiating the associated hypotheses. 



51 



VII. IDENTIFICATION OF EXPERIMENTS 

A total program of space-based experiments for laminar gas-jet diffusion flames is given in 
Table 2. The parameters that will be varied arc the type of fuel, chamber pressure, and chamber 
oxidizer (i.e., 2 /N 2 ratio). The number of tests in the matrix will provide a highly productive 
experimental program. The fundamental effects to be studied are: transient behavior, steady-state 
flame structure, kinetics effects, soot processes, radiation, and the influences of fuel reactivity and 
environmental conditions. 

Hydrocarbons (methane and propane) are selected as fuels because of the available 
background and theory. Methane is selected because of the base of data that already exists. 
Propane is selected to examine a heavier-than-air fuel, and because it is a representative of a wide 
range of higher-molecular-weight hydrocarbons. With these fuels, the effects of reactivity, 
molecular weight, stoichiometry, and diffusivity on the flame structure and behavior, kinetics, 
sooting characteristics, heat release, and radiative transfer will be studied. 

The effects of pressure and oxygen concentration on flames will be studied for a range of 
pressures (0.5, 1.0, and 1.5 atm) and oxygen-nitrogen mixtures (15%, 21%, and 30% O2). Air at 
1.0 atm will be the baseline oxidizing environment. Reduced pressure and/or oxygen 
concentration will be used to approach the near-limit behavior, and to enhance the kinetics effects, 
vary the partial pressure of oxygen, and reduce sooting and radiative loss. High pressure and/or 
high-oxygen-concentration will enhance sooting and subsequent bumoff, in addition to increasing 
the radiative loss and combustion intensity. The combination of various pressures and oxygen 
levels will provide a complete set of data for the effects of environmental conditions on the 
behavior and characteristics of flames in microgravity. 

As described in Section V, certain diagnostics are necessary to provide the data. Of critical 
importance are the temperature map, visualization via cinematography, flame radiation (both overall 
radiation and CO2, H2O, and soot radiation), and chamber pressure. These are identified in 
Table 3 as the "required" diagnostics. In addition to these essential diagnostics, the following are 
highly desired because they significantly enhance the database and provide important information 
on flame characteristics: visualization (both Schlieren and infrared imaging of H2O and/or CO2 
fields), species field, velocity field, acceleration data, soot samples (via thermophoretic sampling), 
and soot concentration (via light absorption/extinction method). These diagnostics are identified in 
Table 3 as "desired." Table 3 also provides the rationale for these diagnostics and application of 



52 



Table 2. Test Matrix. 



OPE RAT INS CONDITIONS 


NO. OF EXPERIMENTS 


SPECIFICS AND REQUIREMENTS 


Fu«i 


2 


Metnane and Propane. 


Fuel Flow Rate 


1 


xed but different no flow race (or 
.(ferent fuels. Whan converted to 
*t«ospherlc condition, 1.0 em'/sec for propana 
and 2.0 ca'/aec (or methane. 


Neziie Radius 


1 


Tapered-tlp nozzle. Innar radius - 0.0»25 em. 


Oxidizer 


} 


Oxygen/nitrogen mixtures; 0,/N, volume ratio • 
0.15/0.15, 0.31/0. 11, and 0.30/0.70. 


Chaxuer Pressure 


2 


P - 0.5, 1.0, and 1.5 atm. 



Total No. of Tlavnas 



Table 3. Diagnostics Requirements. 



VARIABLE 


RECOMMENDED DIAGNOSTICS 


RATIONALE/APPLICATIONS TO HOC EL DEVELOPMENT 


NOTE 


Flam* Structura 
and 
visualization 


Schllaran aystaoi 


To obtain qualitative field of density gradient. 


D 


Cinematography (using. 
movie cajnara or video 
camera) 


Obtaining time-resolved visualization data on 
ignition, flame development, color, luminosity, 
aootlng, and quenching. Flaae shape* can be 
compared directly with model prediction!. 


1 


Temperat ure 


Array of thermocouplea 


Obtaining temperature distributions In the flow 
field. They can be compered directly with the 
model predictions of the thermal field. 


■ 


Comeust Ion 
and Fual 
Da composition 
Prod-cts 


Array of gas sampling 
proDes and sample 
bottles 


Obtaining species distributions which can be 
compared with the aodel prediction of the speclea 
field. 


D 


Cnamcar 
Pressure 


Pressure transducer 


Monitoring any increaae In chamber preaaure due to 
the burning process, which yields estimates of the 
overall heat release rate. 


P. 


Gas Velocity 


S ingle- component laser 
Doppler veloclinetry 
(LDV) , or particle 
Image veloclmetry (PIV) 


Centerllne and radial velocities can be obtained. 
Velocity measurements will be compared directly 
with the model predictions of the flow field. 


PIV: D 

or 
LDV: D 


Fiama Radiation 


Radiomatry 


Isolation of thermal radiation will determine the 
effect of radiation on reduction in temperature 
within the flame. Total infrared radiation, in 
addition to isolated radiation from toot, CO., 
and H,0 can be measured. Model prediction* and 
results of aamplinf for gas-phase species, 
combined with measured radiation, provide the 
contribution* from soot and (a* -phase specie*. 


R 


Gravitational 
Accalaratlon 


Accaleroaeter 


Measurement of gravitational accalaratlon enable* 
ua to interpret the data that could be affected by 
gravitational level* greater than 10"' g. Provide* 
direct eomparlaon* with model prediction* on flame 
characteristics and behavior •■ a function of 
gravitational level. 





Species Imaging 


Infrared imaging 


To map specie* field uainj appropriate optical 
filters with infrared camera. Particularly uaaful 
for K.0 and/or CO, field*. 


D 


Soot Volume 
Fraction and 
Slza 


Light aDsorptlon 


This is * convenient non-intrusive method to Infer 
soot volume fraction and number density (knowing 
soot sit*), (Then combined with radiation data, 
species eessurements, end model predictions, 
provides information on the characteristic* and 
influence of loot in mlcrogravlty. 


D 


Thermopnoretlc Sato; ling 


Soot can be collected and alia distribution can b* 
obtained. However, collected loot need* to be 
removed from the chamber after the teat or * aerie* 
of tests. 


D 


t: Required. O: Daslrad. 



53 



*?»S* 



the data to model development and validation. For any flame of the test matrix of Table 2, the 
"required" diagnostics are essential, and the "desired" diagnostics provide significant data. 

VIII. APPARATUS, CONDITIONS, AND EXPERIMENT TIMELINE 

Experimental conditions, components of apparatus, ranges, requirements, and specifica- 
tions are presented in Table 4 (Experiment Requirements) and Table 5 (Measurement Require- 
ments). The existing chamber for the ground-based microgravity tests is shown in Figure 3, 
which may be used as a guide in designing the new chamber. Assuming that all of the diagnostics 
methods presented in Table 3 are incorporated in the chamber, Table 6 presents a very preliminary 
experiment timeline to obtain a comprehensive set of data on each flame of the test matrix. 

IX. RELATIONSHIP BETWEEN DATA AND MODELING 

The numerical model described in Section DI (0 is in the parabolic form of the partial 
differential equations for mass, momentum, elements, and energy. The effects included are 
diffusion, viscosity, inertia, kinetics, gravity, and radiation, with a shifting-equilibrium chemistry 
model. 

The model is currently being modified to incorporate the effects of: (a) soot formation and 
burn-off, (b) radiation from soot in addition to gas-phase radiation by incorporating the radiative- 
transfer equation in the formulation, (c) thermophoretic effects, which become significant in 
microgravity, (d) quasi-global kinetics, (e) multi-component diffusion in the presence of soot, and 
(0 elliptic effects due to the significance of axial diffusion in microgravity. 

The relationship between data and modeling is presented in Table 3 and in the next section 
on data analysis. 

X. DATA ANALYSIS 

The films from standard flame imaging will be developed and analyzed using a motion- 
picture analyzer. Flame development, height, radius, shape, and color will be obtained as a 
function of time. The data will provide information on the overall characteristics of flames. 
Comparisons of flame shapes and dimensions will be made with the predicted results. 
Observations on color and luminosity will provide information on particulate formation, various 
species, and approach toward steady state. 



54 



Table 4. Experiment Requirements. 



COMPONENT 


GENERAL REQUIREMENTS 


MINIMUM 
REQUIREMENTS 


Experiment 
Cr.amoer 


Cloied, cylindrical chamber. Minimum 
height/diameter - 1.5. Wall reflectivi- 
ty less than 0.05. Burned gas 
recirculation should be minimized at the 
top of the chamber. Recommend mounting 
a mesh at a TBO location near the 
chamber top. Chamber should be sealed 
during the test . 


Minimum volume • 0.035 
m' (critical) ; 0.050 m' 
(desired) . 


Chamber 
Environment 


Initial conditions are as follows. 
Oxlditer: 15% to 30% (by volume) of 
oxygen in nitrogen. Pressure: 0.5 to 
1.5 atra. Before each test, the gat 
temperature should be 295 ♦ /- 5 K 
(measurement accuracy Is ♦ /- 10 K), 
pressure should be P ♦/- 0.02P atm, and 
oxygen concentration should be x% ♦/- 
0.5* on a volume basis (e.g., for the 
15% case, between 14.5% and 15.5%). 
However, these values must be accurately 
known before each test (accuracy: */- 
0.05 %] . 


Oxidizer: 15% to 30% 
of oxygen in nitrogen. 
Pressure: 0.5 to 1-5 
atm. 


Nozzle 


Tapered-tlp, stainless-steel nozzle 
along the centerllne of the cylindrical 
chamber. Minimum distance from wall - 
20 cm. Protruding IS ♦/- 2 cm Inside 
the chamber. Minimum distance from 
nozzle exit to top of the chamber • 35 
cm. The last 4 cm (at least) of the 
nozzle tube (including the tip section) 
must have a constant inner radius. 
Maximum wall thickness is 1 mm for the 
last 4.0 cm of nozzle length. Details 
of the tip Including tapering angle TBO. 


Inner radius - 0.0825 

cm. 


Fuel 


High-quality, commercially available, 
minimum purity 99.9X-pure fuel. 
Methane and propane. Fuel mass-flow 
rate should remain within the 
tolerance throughout the burn. 


Methane and propane. 
Single mass-flow rata 
for each fuel in the 
whole matrix. Volume 
flow rate when 
converted to 
atmospheric condition 
(1 atm) Is 1.0 cc/sec 
for propane and 2.0 
cc/sec for methane, 
tolerance: ♦/- It. 


Combustion 
Time 


The last 10 seconds of the test must 
have no probes Inside the flame (minimum 
a cm away from the nozzle centerllne) . 


Experiment time: 30 - 
60 sec, depending on 
carrier, chamber 
volume, extent of 
diagnostics, and 
particular flame. 


Ignltor 


Ignltor must be in excess of 1200 K be- 
fore the fuel flow starts. Retracting, 
nlchxome, hot-wire ignltor Is 
recommended. Must be at least 4.0 en 
away from the nozzle axis when not 
inside the flame. Ignition occurs 
Immediately following the start of fuel 
flow. Ignltor should remain In the 
flame for approximately 2.0 sec after 
ignition, and then removed in less than 
2.0 tec without extinguishing the flame. 


Same as General 
Requirements. 


Illumination 


Illumination prior to test ia desired to 
be able to locate the nozzle and probes 
on the film. It must be turned off 
before ignition. 


sane aa General 
Requirements. 


Allowable 
Acceleration 
Level* 


Accelerations less than or equal to 10"* 
g are desirable. Frequency should be 
above 10 Hz. 


Saae at General 
Requirements. 



55 



o*<#* 

Of r 



Table 5. Measurement Requirements. 



MEASUREMENT 



GENERAL REQUIREMENTS 



Acceleration 



Charier 
Pressure 



Overall 
Chamber 
Temperature 



Flame 
Radiation 



Soot Size 



Soot volume 
Fraction 



Gas-Phase 
Temperature 
Field 



Acceleration (3 axes) should be recorded continuously 
during the experiment. Data collection rate: minimum of 
30/sec-axls. Acceleromet er in the vicinity of the 
chamber. Range of measurement: 10E-5 to 10E-2 g. One 
axis aligned with the nozzle. Frequency range of 
interest: 0-25 Hi. 



Chamber pressure should be monitored throughout the 
conduct of the experiment at one location on the chamber 
wall. Measurement precision: 1%. Measurement rate: 
1/aec. Reading: 0.05 to 3.0 atm. 



The overall gas-phase temperature should be monitored 
during and between tests. Range: 270 - 1500 K. Collec- 
tion rate: 10/sec. Accuracy: ♦ /- 10 K. Recommend a 
thermocouple located along the nozzle centerllne, inside 
the chamber, and 5 - 10 an away from the chamber top. 
Screen (mesh) is between thermocouple and chamber top. 



Radlatl 
Require 
microme 
being r 
range 
plane 
cm in 
tip. 
Four 
are rec 
wavelen 
one for 



on should be measured thro 
measurements of the entlr 
ters, with at least 95% of 
ecorded. In addition, two 
TBO) should be measured, 
assing through the nozzle 
adlus. Must be able to se< 
ata collection rate: 20/se 
ide-vlew-angle, thermopile 
ommended: one for overall 
gth range), one for a wate 
a CO., band (TR) TBD, and one 



ughout the test, 
range 0.3 to 9.0 

Incident radiation 
discrete bands in the IR 
Field of view at the 
centerllne: circle of 20 
e 5 cm below the nozzle 
c. Precision: 1%. 

detector radiometers 
radiation (entire 
r vapor band (IR) TBO, 

for soot (IR) TBD. 



Soot samp 
in the fl 
nozzle ax 
Collectio 
time towa 
mechanism 
nozzle ce 
therraopho 
collectio 
at least 
in the fl 
flame in 
removed f 
II) and a 
specifics 



les must 
ame cove 
is) and 
n durati 
rd the e 
is used 
nterllne 
retic-sa 
n mechan 
two grid 
ame, up 
no more 
rom the 
fter ret 
, and me 



be co 
ring a 
15 cm 
on: le 
nd of 
it s 

when 
mpling 
ism sh 
s per 
to the 
than 3 
chambe 
urn to 
chan is 



llecte 

regio 
vert ic 
ss tha 
the tt 
hould 
not In 

mecha 
ould h 
arm. 

nozzl 

msec 

at t 

Earth 
m TBD. 



d from 

n 2 cm 

al (a 

n 30 m 

st. I 

e at 

the f 

nism 1 

ave at 

Mechan 

e cent 

Not 

he end 

(Phas 



at least fo 

horizontal 

ong the nozz 

sec at a spe 

f a flame-pe 
least 6 cm a 
lame. A ret 
s recommends 

least four 
ism actlvatl 
erllne, and 
e that sampl 

of the test 
el). Locat 



ur points 
(from the 
le axis) . 
cifled 
rturblng 
way from 
ractlng, 
d. The 
arms and 
on: going 
out of the 
es must be 

(Phase 
Ions, 



Measurement of soot volume fraction in a 10-cm-dlameter 
field and centered 10 an above the nozzle tip at the 
centerllne of the nozzle. Light absorption/extinction 
method is recommended. A beam of light, passing through 
the flame at a specified time, and collected with a 
detector. Requires (preferably! a monochromatic beam. 
Specifics TBD. Detector collection rate: 20/sec. 
Detector specifics: TBD. 



Radial temperature distributions at 6-» TBD axial 
locations must be measured. Upper limit: 1S00 K. 
Sampling rate: 10/sec. Measurement accuracy: ♦ /- 10 K. 
Measurements at TBD locations in a plane passing through 
the nozzle axis. Region of interest Is 5 cm horizontally 
outward from the nozzle axis and 20 cm vertical (along 
nozzle axis) . It is recommended to use a rake of 6-8 
thermocouples on a single line parallel to the nozzle 
axis which can be moved In the radial direction (to 4-5 
cm from the axis) at a controlled rate (1-2 tun/ sec ) . 
Separation between thermocouples: TBD (2-4 cm). 
Tolerance on the location of the TC tip is ♦/- 0.5 ami. 
Type, wire diameter, and bead size: TBD. It is preferred 
to move the rake in a direction perpendicular to the 
thermocouple axis; If not possible, then along the axis. 
There are other possibilities, e.g., turning the 
mechanism along an arc going In the flame and continuing 
out of the flame. Controlling factors for the selection 
of mechanism are: experiment duration, minimum flame 
perturbation, rata of movement of the mechanism, and 
local flame quenching by thermocouple arms. 



MINIMUM 
REQUIREMENTS 



Same as 

Ceneral 
Requirements. 



Same as 
General 
Requirements. 



Same as 
General 
Requirements. 



Same as 
General 
Requirements. 



For a selected 
number of 
tests (TBD) . 



Same as 
General 
Requirements. 



Same aa 
General 
Requirements. 



56 









^ 



Table 5. 



Cont, 



MEASUREMENT 


GENERAL REQUIREMENTS 


MINIMUM 
REOOIRE MINTS 


Concentration 
of Species 


Concentration measurements of CO,, H,0, CO, H, , 
2f N a , formaldehyde, unburned hydrocarbons, 
etc. Kill analyze the samples after tests are 
done. Location: TBD; separation: TBD. Rake of 
gas-sampling probes and associated sample 
bottles are recommended. Quenching quartz 
mlcroprobes. At a specified time, and 
preferably, for a specified period during the 
test, they will remove gas samples from the 
flow field by suction Into vacuum teflon-lined 
sampling bottles of approximately 10 cc In 
volume. A rake of 3x3 Is recommended. Region 
of Interest Is S cm radially outward from, and 
20 an vertical along, the nozzle axis. One 
extreme column to withdraw samples at the 
centerllne. Separations and locations TBD. 
Should not affect or Interfere with temperature 
measurement. Probes Inside the flame for a 
specified time only. 


As a minimum, 
three sampling 
probes: one at 3 
cm away from the 
nozzle centerllne 
and 5 cm above the 
nozzle tip; one 
above and one 
below the flame 
tip (locations 
TBD) at the 
centerllne. All 
three measurements 
simultaneously at 
a specified time. 


Vei oclt y 
Measurement 


Gas-phase velocity measurement In and around 
the flame. Range of velocities: 0.1 to 500 
cm/sec. Point measurements are preferred If 
velocity can be measured at several (TBD) 
points; otherwise, field measurements. Point 
measurements: axial velocities at TBO locations 
In the flame region In a plane passing through 
the nozzle axis. Field measurements: 2-D 
velocity vector field. In a plane passing 
through the nozzle axis. The entire region of 
Interest Is IS cm horizontal by 25 cm vertical, 
starting at 0.2 - 0.4 cm above the nozzle tip. 
Spatial resolution: 1 mm. Measurement accura- 
cy: ♦ /- 5» of reading. Sampling rate: 30/sec. 


Same as General 
Requirements. 
Single-componenet 
laser Doppler 
veloclmetry (LDV) 
is recommended if 
point measure- 
ments; particle 
image veloclmetry 
(PIV) if field 
measurements. 


Flame Imaging 
(Standard) 


Side view of the flame to be imaged in color. 
Field of view: 15 cm horizontal by 25 cm verti- 
cal (must see 2 cm of the nozzle tip) . A 
timing system should be provided to record 
accurate time Intervals on the film (approx. 
0.1 sec). If a mirror is used, should not pro- 
duce double-imaging on the film. Framing rata 
of 30 fps is desired. Camera should see the 
perturbation of the flame by thermocouples and 
sampling probes (approx. 90 degrees to probe 
arms) . Resolution: 0.5 mm. 


Same as General 
Requirements. 


Infrared 
Imaging for 
Species 


Mater vapor and/or carbon-dioxide imaging. 
Field of view: 20 cm horizontal by 30 cm 
vertical. View must see 2 an of the nozzle 
tip. Focus on nozzle axis. Resolution: 0.5 
mm. Imaging throughout the test. Depth of 
field: 2-4 cm. Infrared camera is recommended. 
Filters (wavelengths of interest to be 
specified) are needed to identify the CO, and 
H,0 fields. Framing rate: 30 fps. 


Same as General 
Requirements. 


Density- 
Gradient 
Field 
Imaging 


Imaging of density gradient to see the entire 
flame and region above it. Minimum field of 
view is a circle IS cm in diameter, and 
centered on the flame axis; center of the 
circle is 7.0 cm from the nozzle tip. Desired 
resolution: 0.S am. Schlleren system Is 
recommended. Frame rate: 30 fps. Either color 
or B/H; color preferred. 


Same as General 
Requirements. 



57 



*0& 



Table 6. Preliminary Timeline for Each Test 
(Times are in Seconds) 



t < o : Thermocouples are at nozzle centeriine. Ignitor is above the nozzle tip. 

t - : Start recording: accelerometer, movie camera, thermocouples, pressure transducer, infrared 
camera, radiometers. Ignitor heat-up starts. Fuel-flow starts. 

t ■ 1-2. Ignitor is moved away from nozzle centeriine. 

t - 10: Thermocouple rake starts to move away (rate: 1.5 mm/sec). Light absorption/extinction starts. 

t - 30: Light absorption/extinction stops. Schlieren starts. Soot-sampling mechanism starts to move 
toward the flame. 

t . 34: Soot-sampling mechanism stops near the flame (at least 5-6 cm away from nozzle centeriine). 
t - 35: Soot sampling is conducted. 

t - 36: Soot-sampling mechanism starts to move toward its rest position. Gas-sampling mechanism 
starts to move toward and into the flame. 

t . 40: Thermocouple rake stops moving. Soot-sampling mechanism arrives at its rest position. Gas- 
sampling probes are now inside the flame. Gas sampling starts. 

t » 42: Gas sampling stops. Gas-sampling probes start to move out of the flame and toward their rest 
position. 

t - 46: Sampling probes get to their rest position. Schlieren system stops, 
t - 55: Particle-Image Velocimetry (PIV) system is activated, 
t - 65: Fuel flow stops. 

t * 70: PIV system is turned off. Infrared and cinematography cameras stop. Thermocouples, 

accelerometer, radiometers, and pressure transducer stop recording. Experiment is ended. 



58 



Acceleration data will be reduced to obtain the flame behavior as a function of any small 
variations in gravity level. These data combined with the results of film analysis will provide 
information on the response of the flame to any g-jitter effects. These acceleration data will be 
used as input to the computer programs for prediction of flame characteristics. 

The gas samples withdrawn from various locations inside and outside the flame will be 
analyzed. Concentrations of O2, CO, CO2. N 2 , H2, and unbumed hydrocarbons will be obtained, 
which provide species distributions. The data will be compared with the predicted values of 
species field, and will specifically provide information on any potential escape of unbumed fuel 
through the tip of the open flames. 

Radiation data will be analyzed to obtain heat loss as a function of time. Any approach 
toward steady state will be compared with the results of flame visualization data. The measured 
overall radiative loss combined with the data from discrete-radiation measurements on CO2. H2O, 
and soot will provide a complete breakdown of radiation contribution from these stable products 
and from particulates. The radiation model in the numerical code will be validated using these 
measurements. 

The data from temperature-field measurements will provide the distribution of temperature 
in both axial and radial directions. In addition, flame temperature and location, far-field 
temperature as a function of time, and the overall gas-phase temperature will be obtained. These 
data will help quantify tip-quenching due to the presence of soot. When combined with the 
radiation data, species measurements, and the predicted species field, the temperature data will 
provide a complete map of effective emissivitics and absorptivities. 

The chamber-pressure data will be used to deduce the overall heat-release rate and 
combustion efficiency. 

The data from velocity measurements will be analyzed, and a complete map of the velocity 
field will be constructed. The tip-opening phenomena in microgravity flames will be better 
understood through the mapping of velocity field at the flame tip. These data together with the 
theoretical model will quantify the thermophoretic effects. 

The collected soot samples will be analyzed to obtain soot size and overall soot distribution 
in the flame. Once the soot size and soot concentration are measured, the input from soot contribu- 
tion in the submodels for radiation and soot formation/agglomeration/bum-off will be quantified. 



59 



The data from concentration measurements for soot will be analyzed to construct a map of 
soot concentration, and to better characterize the effects of thermophoresis, soot blockage and 
coring effects, and soot quenching, which contribute to flame extinction at the tip. When 
combined with the soot-collection data, a complete set of information on soot size, number density, 
and cor entration will be obtained. 

The infrared- imaging data for CO2 and/or H2O will provide the local relative intensities of 
contributions from stable combustion products. The field of products will be monitored through- 
out the test which will help to define the location of the flame and the extent of the field during the 
experiment. 

The films of the density-gradient field imaging will be analyzed to obtain information on 
flame structure, the fields of temperature and density, and help to define the perturbations caused 
by the presence of probes in the flame. 

Theoretical predictions will be used to support the analysis and interpretation of the data- 
base developed from these experiments. The results of individual measurements, when combined, 
provide both quantitative and qualitative information to validate and improve the theoretical 

analyses. 



60 



REFERENCES 



1 . Edelman, R. B. and Bahadori, M. Y. (1986) Effects of buoyancy on gas jet diffusion flames: 
Experiment and theory. Acta Astronautica 13, No. 11/12, 681-688. 

2. Burke, S. P. and Schumann, T. E. W. (1928) Diffusion flames. Ind. Eng. Chem. 20, 
998-1004. 

3. Astavin, V. S. and Ryazantsev, Yu. S. (1979) Temperature and concentration distribution in 
the reaction region of parallel flows of unmixed reactants. Fluid Dynamics 14, 274-279; 
Translated from Izvestiva Akademii Nauk SSSR - Mekhanika Zhidkosti i Gaza 14, 139-146. 

4. Bahadori, M. Y., Li, C-P., and Penner, S. S. (1986) Two adjacent, coupled laminar 
diffusion flames With cylindrical symmetry. Progress in Astronautics and Aeronautics. 
Vol. 105 - Dynamics of Reactive Systems. Pan I: Flames and Configurations (J. R. 
Bowen, J.-C. Leyer and R. I. Soloukhin, Eds.), pp. 192-207. American Institute of 
Aeronautics and Astronautics, New York. 

5. Penner, S. S., Bahadori, M. Y., and Kennedy, E. M. (1984) Laminar diffusion flames with 
cylindrical symmetry, arbitrary values of diffusion coefficients and inlet velocities, and 
chemical reactions in the approach streams. Progress in Astronautics and Aeronautics. 
Vol. 95 - D ynamics of Flames and Reactive Systems (J. R. Bowen, N. Manson, A. K. 
Oppenheim, and R. I. Soloukhin, Eds.), pp. 261-292. American Institute of Aeronautics 
and Astronautics, New York. 

6. Bahadori, M. Y. (1990) An analytical solution for transient, cylindrically symmetric laminar 
diffusion flames in the absence of buoyancy. Spring Technical Meeting (Central States 
Section) of The Combustion Institute, Cincinnati, Ohio. 

7. Beckstead, M. W. (1981) A model for solid propellant combustion. Eighteenth Symposium 
(International) on Combustion, pp. 175-185. The Combustion Institute, Pittsburgh, PA; see 
also Derr, R. L., Beckstead, M. W, and Cohen, N. S. (1969) Combustion tailoring criteria 
for solid propellants. AFRPL-TR-69-190 (835-F), Redlands, CA. 

8. Mitchell, R. E., Sarofim, A. F., and Clomburg, L. A. (1980) Experimental and numerical 
investigation of confined laminar diffusion flames. Combust. Flame 37, 227-244. 

9. Edelman, R. B., Fortune, 0., and Weilerstein, G. (1973) Analytical study of gravity effects 
on laminar diffusion flames. NASA CR- 120921, 136 pp. 

10. Edelman R. B., Fortune, O. F., Weilerstein, G., Cochran, T. H., and Haggard, JR., J. B. 
(1973) An analytical and experimental investigation of gravity effects upon laminar gas jet 
diffusion flames. Fourteenth Symposium rinternationan on Combustion, pp. 399-412. The 
Combustion Institute, Pittsburgh, PA. 

11. Bahadori, M. Y., Edelman, R. B., Stocker, D. P., and Olson, S. L. (1990) Ignition and 
behavior of laminar gas-jet diffusion flames in microgravity. AIAA J. 28, 236-244. 

12. Klajn, M. and Oppenheim, A.K. (1982) Influence of exothermicity on the shape of a 
diffusion flame. Nineteenth Symposium flnternationan on Combustion, pp. 223-235. The 
Combustion Institute, Pittsburgh, PA. 



61 



1 3 Flower W. L. and Bowman, C. T. (1983) Measurements of the structure of sooting laminar 
' diffusion flames at variable pressure. Report SAND83-8776 Sandia National Laboratories 
Livermore, CA 94550; paper Western States Section of the Combustion Institute, Wbb/CI 



83-56. 



14 Cochran, T. H. and Masica, W. J. (1970) Effects of gravity on laminar gas jet diffusion 
flames. NASA TN D-5872. 

15 Cochran, T. H. (1972) Experimental investigation of laminar gas jet diffusion flames in zero 
gravity. NASA TN D-6523, 26 pp. 

16. Haggard, J. B. and Cochran, T. H. (1973) Hydrogen and hydrocarbon diffusion flames in a 
weightless environment. NASA TN D-7165, 28 pp. 

17. Haggard, J. B. (1981) Forced and natural convection in laminar jet diffusion flames. NASA 
Technical Paper No. 1 84 1 . 2 1 pp. 

1 8 Bahadori, M. Y. and Stocker, D. P. (1989) Oxygen-concentration effects on microgravity 
laminar methane and propane diffusion flames. Fall Technical Meeting (Eastern Secoon) of 
The Combustion Institute, Albany, New York. 

19 Bahadori M Y., Stocker, D. P., and Edelman, R. B. (1990) Effects of pressure on 
micwavity hydrocarbon diffusion flames. Paper AIAA 90-0651, AIAA 28th Aerospace 
Sciences Meeting, Reno, Nevada. 

20. Bahadori, M. Y., Edelman, R. B., Sotos, R. G., and Stocker, D . P. W^™^^™ 
gas-jet diffusion flames in microgravity environments. Paper AIAA 91-0719, AIAA 2Vth 
Aerospace Sciences Meeting, Reno, Nevada. 

2 1 Bahadori, M. Y., Edelman, R. B., Sotos, R. G., and Stocker , D . P. (1990) Measurements 
of temperature for microgravity laminar diffusion flames. Fall Technical Meeting (Eastern 
Section) of The Combustion Institute, Orlando, Florida. 

22 Altenkirch, R. A., Eichhom, R., Hsu, N. N., Brancic, A. B., and Cerallos, N. E. (1976) 
" Characteristics of laminar gas jet diffusion flames under ** ^™? { ^"^JtZ'Sn 

fij*t,-nrh Svmnorinm flmeraatio nan on Cnmhnstion. DP. 1165-1174. The Combustion 

Institute, Pittsburgh, PA. 

it. Kimura I and Ukawa H (1961) A study of the combustion of laminar fuel jets. iaBAn 
23 ' S^ifiS^to 27. 736-746; translated as NASA Technical Translate, 
NASA TT F-13, 459, February 1971. 

24 Stocker D P (1990) Size and shape of laminar Burke- Schumann diffusion flames in 
mkroSaviiy. Spring Technical Meeting (Central States Section) of The Combustion 
Institute, Cincinnati, Ohio. 



62 



Section 19 



'EFFECTS OF BUOYANCY ON GAS-JET DIFFUSION FLAMES" 



-Concept Design- 
(Space Experiment) 



M. Y. Bahadori and R. B. Edelman 



December 1991 



Effects of Buoyancy on Laminar 
Gas-Jet Diffusion Flames 



(Space Experiment) 

- Science Requirements and Concept Design - 
Contract NAS3-22822 

M. Yousef Bahadori and Raymond B. Edelman 

Science Applications Internationa! Corporation 

Thermal Sciences Division 

21151 Western Avenue, Torrance, California 90501 



Presented at: 

NASA Lewis Research Center 

Space Experiments Division 

Cleveland, OH 44135 



V. 



December 11, 1991 



Introduction 

and 
Background 



5AIC 



A 



SAIC 



Laminar Gas-Jet Diffusion Flames 



>v 




riAM • TA(UIA1»H 




1-g 



0-g 



Jet diffusion flames embody nearly all the mechanisms found in natural 
fires relevant to fire safety and to controlled combustion processes 

Simple, basic, and easy to control 

Need better understanding 

A step toward understanding the more complex turbulent diffusion flame 



SAIC 



>v 



r 



Laminar Gas-Jet Diffusion Flames (Cont.) 



lirmiw 
llimts 



Iijnuiion 
rig ion 



fully dtvlloptd 
hiivultni llimts 




Volume flow rite (or Reynolds number) 



V. 



Understanding laminar flames is basic to the physics of the 
broader, more general turbulent combustion process 

The transition process to turbulent flames is governed, in part, 
by the characteristics of laminar flames 

Laminar flames are inherent in the detailed structure of 
turbulent flames from the molecular point of view 



SMC. 



r 



Approach 



• Isolate the effects of buoyancy by eliminating gravity 

• Experimental (2.2-sec. tower, 5.18-sec. facility, KC-135 aircraft, and 
space experiment) 

• Theoretical (analytical and numerical modeling) 



Results applicable to the fundamental understanding of combustion in 
space and earth environments 



5AIC 



r " \ 

Previous Research - 2.2-Second Drop-Tower Experiments* 



Hydrocarbon flames were ignited in 
normal gravity and then subjected to 
microgravity condition 

• Some flames extinguished 

• Some were in transition between 
extinguishment and steady-state 
(lack of enough time) 

• Others reached (an apparent) 
steady-state with flame-height 
fluctuations 




Extinguished microgravity methane diffusion llame 
Nozzle radius « 0.186 cm 
Flow rate » 2.8 cm3/sec 
Flow velocity - 25.44 cm/sec 
Reynolds no. - 28.4 



I.t 


; 


(•> 


Deported 

ExtkicHon 
> 

: 
• r 

• 


7.S 






! •' 


7.1 






• 


1 7 




■ i 


! iw-f 
■ i ■ . i i 



I.I «.t 



V 



Cochran, Masica, Haggard, Edelman (1970-1974). 



5. J 
MC 



5NL. 



Previous Research - Theoretical Analysis 




- 


".0. 


co,. : 


'•-'7 




■t " 


- 


, , 1 


* l 






f ipe rl« 


tnlt\ <J«t»*' 


Norai«l . 


9 


r. (c-l 


o 




0.0M 
0.0«J5 


a 




Oil] 







0.186 


o 

V 




o.m 
o.««? 



PirtUJ oildltlon 
theory 



M,0, CO M,0, C(») 

O -D 



r, (c«> 



0.051 
0,M?J 



Sol t* ijMbols denote 
I«ro-f 

•Cochran (1972) 

' Cochran and MaslcaU970) 



Soli* lyabols ornott 
eiperto-ewtil oili* 



Dot«* 

Q • 1 .08 ca'/tcc 
»«,• «0.l 

r, • 0.051 m 



ANALYSIS 



Used 



Steady flow 

Axisymmetric, 2-D conservation equations 
lor mass, momentum, and energy 
Parabolic form of governing equations 
Infinitely fast reactions with shifting equilibrium 
chemistry, in addition to partial oxidation 
Mutticomponent diffusion 



Required 

Kinetic effects in the presence of soot processes 
Flame radiation 

Transient effects, ignition, and extinction 
Axial diffusion 



SAIC. 



Tl*>w M*« minrm l~n— r~n — ' i 



r 



Current Ground-Based Program 

• 2.2-Second Drop-Tower Tests 

• 5.18-Secbnd Microgravity Tests 

• KC-1 35 Tests 

• Theoretical Analyses 



V. 



SMC. 

i, >i i — 



Background Summary 



Buoyancy significantly alters the flame structure 

Pressure and oxygen concentration have a more dramatic effect in 
microgravity than in normal gravity 

Prolific sooting results from absence of buoyancy, and is strongly 
pressure- and oxygen-dependent 

Radiation and temperature data show that longer test times are 
needed to substantiate the observed behavior of these flames at 
steady-state 

Chemical kinetics - radiation - transient interactions are important in 
reduced-gravity environments 

Extinction limits may be different in microgravity and normal gravity 



SAIC 



Ju<i.i * u m 1 1 1 l—i— * rn ri w rr r 



Transition from Ground-Based to Space Program 



Past Effort 
(1970-1974) 



Observations and 
predictions from: 
hypotheses 
•2.2 sec. tests 
•Modeling 



Current Ground-Based 
Program (1987-1991) 



Results have shown: 
•Lack of quantitative 

data 
•Need for longer 

test times 
•Improvement In models 



Data, observations, and 
predictions from: 

•2.2 sec. tests 
•5.18 sec. tests 
•KC-135 tests 
• Modeling 



T 



Results have shown: 
•Significant effects of pressure 

fuel, and oxid zer in 

microgravity 
•Need for longer test times 
•Need for higher-fidelity data 
•More advanced diagnostics 
•Improvement in models 



Proposed Space Program 
(1992- ) 



•Overall and Specific Objectives 
•Hypotheses 

•Well-defined test matrix 
•"Required" diagnostics 
•"Desired" diagnostics 
•Experimental duration 

and timeline 
•Supporting ground-based effort 
•Associated modeling 
•Post-flight data analysis 



7 

win 



Results wifl significantly enhance 
our understanding of gas-jet 
diffusion flames In microgravity 



Is Space Necessary? 



Requirement 


Criticality 






G-level has to be <10-3 


Critical 


Time Requirement: 

• Achieve steady state (>5sec.) 
•Obtain quantitative data (conventional 
or advanced diagnostics -60sec.) 


Critical 


Quantitative data via advanced diagnostics 


Desirable 



r 



"\ 



Current Ground-Based Program 



V. 



">v 



r 



Current Ground-Based Program - 2.2-Second 
Drop-Tower Experiments 

Approximately 100 tests were conducted 

Flames were ignited in microgravity (using a newly developed spark 

ignitor) 

Flame extinction was not observed (unlike previous works), once the 
flames were ignited in low-g and the photographic technique was 
improved 

Studied methane and propane flames under various environmental 
conditions of pressure (0.5-1 .5 atm) and oxygen concentration 
(18%-50%O 2 ) 

Defined and revised the test matrix for the 5.18-second microgravity 
experiments, based on findings from the 2.2-second tests 

Observed some new and unique characteristics of microgravity 
flames (e.g., transient effects, blue flames, prolific sooting, 
tip-opening) 

Representative results will follow 



SMC, 



Current Program - 5. 18 Second Microgravity Tests 



Approximately 45 successful tests were conducted 

Flames were ignited in microgravity (originally with a spark ignitor, 
and later with a more reliable, retracting hot-wire ignitor) 

Studied methane and propane flames in different pressure (0.5-1.5 
atm) and oxygen-concentration (15%-30%O 2 ) environments 

Employed flame visualization using movie camera 

Measured flame radiation 

Obtained temperature distributions at various locations 

.Observed new and unexpected phenomena 

Defined the test matrix for KC-135 experiments 
Species distributions via gas sampling will be obtained 
Representative results will follow 



SAIC 



Current Ground-Based Program - KC-135 Tests 

• Testing has begun (first week of flights: visualization only, oxidyzing 
environment: air at 1 atm) 

• Test matrix is selected based on 5.18-second microgravity results 

• Flame visualization via video camera and movie camera 

• Chamber pressure has been recorded 

• Temperature distributions will be obtained 

• Flame radiation will be measured 

• Acceleration data will be recorded and compared with flame behavior 

• Both attached (~10 2 g> and free-float (-10 J g) tests 
Results from the first flight will follow 



■SBC' 



Ground-Based Test Matrix • Current Program 

5.1 8-Second Microgravity Tests 



Fuel 



Oxidizer 



Nozzle radius 



Fuel flow rate 



Pressure 



Fuel 



Oxidizer 



Nozzle radius 



Fuel flow rate 



Pressure 



CH 4 and C 3 H 8 



18% p2, 21% Q 2 , and 30% Q 2 in N 2 



0.0825 cm 



Methane: 1 .0, 2.0, and 3.0 cc/s 
Propane: 0.5. 1 .0. and 1 .5 cc/s 



0.5, 1.0. and 1.5 atm 



Total: 



KC-135 Tests 



CH 4 and C 3 H 8 



15% Q 2 , 21% O;, and 30% Q 2 in N 2 
0.0825 cm 



Methane: 2.0 cc/s; propane: 1.0 cc/s 



0.5, 1 .0, and 1 .5 atm 



Total: 



54 



18 



fSMC 



Hardware - 5. 18-Second and KC-135 Tests 



cas sampling prools 
wd thermocouples 




fUEL HOllll 
THERMOCOUPLES 



SPAM ELECTRODE 
&AS SAHPLIIIC PROIES 



NOV 1 1 CAMERA 




MOVIE CAMERA 



SPAM ELECTRODE 



Radiometer 

Rake of 9 thermocouples 

Rake of 9 sampling probes 

Pressure transducer 

Movie camera and video camera 

Accelerometer 



V. 



SAIC 



r 



>v 



Observations from Current Ground-Based 
Gas-Jet Experiments 



Under reduced-gravity conditions, flames exhibit distinctive characteristics 
relative to normal-gravity flames 



Flame flicker disappears 

Flames become taller and wider 

Take much longer time to reach steady-state 

Drastically change in color and luminosity 

Removal of buoyancy results in prolific sooting 

Tip-opening and soot escape is observed for higher hydrocarbons 

Show significant radiative loss 

Results have helped to clarify what was previously believed to be 
quenching of gas-jet diffusion flames in microgravity 



SMC 



Results from Current 
Ground-Based Program 



5AIC 



Ignition Behavior and Flame Development 




Frame-by-frame analysis of t microgravity propane flame with nozzle radius ■ 0.082S cm and volume flow rite 

- 1 .0 cmVs. The different observed colon are as follows: OBright white; QWhite with Irak yellow; 

B White with more noticeable yellow; BBlue: BDark blue; BViolet-pink; BMixed bhje/violet; PYeD ow; 

B Orange; B Cherry red; BDull red. fainting: Q Color at the base, starting with dark at the bottom, becoming 

dark blue/pink, then pink/blue, then pink/orange, then orange/yellow, and then yellow/white toward the center of 

the flame; ■ Bright white spark with bright blue at its boundary. Scalei m a Icm. 



' Reproduced from Bahadori. et al., 1990 [7] 



V. 



SMC 



Effects of Oxygen Concentration* 




Normal gravity and micrognvuy hwaw di/fiif ion flamea of methane and propane, 
burn in | in quiescent enwonmanu at I aun with dl Hasan l oxygen-nitrogen 
mixtures. The two nooJei haw a lapcred-lip reditu of 0.0*8 and 0.074 cm. The 
various colors indicated ui the diagram art u follows: B (blue). D (da/k). O 
(orange). P (pink). R (red). W (while). Y (yellow), and DR (duJl red). The ban 
iho* the range of normal- privity flame flicker (0. The coloring nay be ilightijr 
ofTdue to the force -processing of ffbru for low-oxygen and 21%-Ojearnronmenu. 
The rums for the flames in XMfc-Oj eavvonrnenu were not force-processed. The 
tow-oiygen-conceiumion Dame of propane in normal gravity shows pockets of 
fUme leaving (he fUckcring pan. Scale; •— — « * 1 cm. 



Reproduced from Bahadori and Stacker, 1989 [4] 



Most of the low-0 2 flames are blue and cool 

Apparently soot-free 

Contrary to what was expected 

May be able to control the soot process 
by the environment 

Open tip {underventilated-type behavior) is 
attributed to: 

Enhanced radiative loss due to larger 
flame size 

Extensive soot and cooling at the tip 

Thermophoretic effects, which become 
more important in microgravity 

Even at 50% 2 , open tip and massive soot 
escape through the tip were observed in 
microgravity 

Soot quenching may not be avoided in 
microgravity by merely increasing the 
oxygen level 



5AIC. 



j 



r 



Effects of Oxygen Concentration (Cont.)' 



3. 

i i 




i ' ' ' ■ 



U » U 1* M » M U 



u 


: 




rmmm r 




u 


- 


^_* 


o . 
o ■ 




1" 


i — *— i 1— ( t-..i— 


—ii — 


A A 


■ W «/• 


3 • 

J. 

i 









1* ti to u 



Methane 



Propane 



o The difference between normal-gravity flame heights decreases as oxygen 
concentration increases 

o However, color, luminosity, and soot formation are different 

o Quantitative data is helping to characterize these processes including extinction 
limits in terms of oxygen concentration 



* Reproduced from Bahadori and Stocker, 1989 (4) 



sm: 



r 



"\ 



Effects of Pressure* 











Increase in pressure has a 
significant effect on tip opening and 
increased soot in microgravity 

Flame radius decreases with 
increase in pressure 

Blue-base region (soot inception 
zone) decreases in length with 
increase in pressure 

At low pressures, less-efficient 
burning and broader flame zone 
exist due to the kinetics effects 



Nomd gravity ml inaofn'ity fames of pnaaac ia iir with nouk diimncr • 
1 5 cm tnd fuel Reynolds number • 91 (sued cm the nozzle ndus). TV vinous 
colon observed tre is follows: B (blue). BB (Wi|hi Muc). O (d»rk). DB (dtst 
blue). DP (dirk pink). OR (dirk fid). O (araifc). P (pink). R (red). W (white). 
aadY(jcl)ow). The range of flictcT (f) is liso » fkr ii n ) ton " 

scale: - 1 cm; reproduced froca (I). 



V 



Reproduced from Bahadori, et al.. 1990 [5] 



SAIC 



Samwrntt^mmam i mimmmC tM i t ama'' 



Effects of Pressure (Cont.)' 









* • 1 S *%w 


10 






^SjT* * *** 


I . 






^S^ **^* • ■• 








i 
i • 

! . 

I 


■ 










-» ' 


■ ' 

















ts 

i 










■* ■ 


x«S 


• 
t 
i i 












***>S^ 


I 












l!^-*""'*^ 






* 




1 


















-»— 


i 









*r+\**r* (•<*) 




Hicp-oqravlty flwe nfloht vs. fuel 
Reynolds number for different pressures. 
Th* lines connect the cUU point! . Note 
thet the flint htloht it torgtst it 1.5 
iw *nd MMUflSt it 1.0 its. 



Ntcroorivity fl*"«e height vs. pressure 
for different full Reynolds nwicer. The 
line* connect the d«ti points. Note 
thit it eich Seynolds number, there 
e*'sti a pre i lure betwMten 05 ind 1.0 
iti it «hich the fliae \t shortest. 



Fl«»e-ttp dloaeter vs. full Reynolds 
Nuaoer it different pressures for Micro- 
|riv1tv floats. The broken Hue indi- 
cates uncertainty in connecting the diti 
points due to the lick of Information on 
the uuUetl Reynolds niottr for which 
the tip opens. 



For the range of pressures studied, there exists an intermediate pressure at which flame height is 

minimized 

Competition between the different pressure-dependent gas-phase combustion rates and soot 
processes may be responsible for this non-monotonic behavior 

Flame-tip diameter has an asymptotic type of convergence with increasing Reynolds number 

Regardless of pressure (and hence, availability of oxygen), flame tip has a fixed diameter 
beyond a certain Reynolds number 



* Reproduced from Bahadori, et al., 1990 [5] 



5AIC. 



"N 



Effects of Pressure (Cont.)' 





Soot Quench length (red region at the 
flame tip) vi . fuel Reynolds nu»6e»" as a 
function of pressure. The solid linei 
connect tft* data points. 



BUe-bait height vs. fuel leyno'ds nuater 
at different pressures fo' noraal-gr*vi ty 
fines (open synbols end dashed lines) 
a«d "iCrogravUy flaaes (ctoted SyafcolS 
and solid lines). Whenever the normal- 
gravity and "icrogrjvity data co<«Ode. 
only the »icro9rivity iy«*ol is tiled. 
The lines connect the data points. 



Soot quench length (at the tip) increases with both pressure and 
Reynolds number, indicating enhanced soot formation 

As pressure increases, both microgravity and normal-gravity 
soot-inception regions (i.e., blue-base region) decrease in length 
with little dependence on flow rate 



* Reproduced from Bahadori, et al., 1990 [5] 



SAIC 



"\ 



Flame Radiation - Comparison Between 
Normal Gravity and Microgravity 



Thcrnjl 

Kidltllon 

<»V) 



12 
































/ 


0-G 










A 


/ 








V 




/v 






1-G 




^ 





/ p. 













Tlat (mc) 



Propane flames burning in quiescent air at 1 atm; nozzle radius - 0.0825 cm; fuel-flow rate - 1 .5 
cm 3 /sec; recent 5.18-second microgravity tests 

Radiative loss is constant for normal-gravity flames throughout the test; 
this is largely due to 

• Instantaneous removal of the hot combustion products under the 
influence of buoyancy 

• Immediate flame establishment following ignition 

In 5-second microgravity environments, a continuous development of the 
flame is observed 

• Radiation increases continuously; more than 5 seconds is needed to 
reach steady-state (verified both visually by examining the films and 
through radiation data) 

• This is due to the continuous accumulation and slow transDortpf the 
combustion products in the vicinity of the flame JSot/LT 



J 



Flame Radiation - Effects of Pressure 



Normal Gravity 



Microgravity 































































. /— - 













P - 1 atm 



211 0, 


































































. Ka, 




_/•— -*— 


■ 


—* •» 1 

















































v 


/' 








\ 


Z 










\ 


« 


























































~—V^ 






-vy 






I ^ 



P ■ 1/2 atm o 
211 0, 1 



t (sec) t (sec) 

Methane flames burning in quiescent air; nozzle radius - 0.0825 cm; fuel-flow 
rate - 3.0 cm 3 /sec; recent 5.18-second microgravity tests 



Pressure has a significant effect on radiative loss in microgravity 
Normal-gravity flames do not show this magnitude of difference 



SPJC. 



nMpWMMC 



"\ 



r 



Flame Radiation - Effects of Oxygen Concentration 



Propane 
1 .0 atm 







• Oxygen concentration has a significant effect on radiative loss in 
microgravity 

• Normal-gravity flames do not show this effect due to the presence 
of buoyancy 

SMC. 



Flame Radiation - Effects of Fuel-Flow Rate* 



1 



£ 3.0 




0.5 1.0 1.5 

Voluat Flo. lilt (oi 3 /itc) 

Radiance as a function of fuel 
volume-flow rale for propane flames 
burning in quiescent air at 1 atm. The 
radiance data are the average values 
between 4 0-sec. and 5.0-sec. after 
ignition in the 5.18-Second Zero 
Gravity Facility 



* Reproduced from Bahadori, et al., 1991 [10] 



Fuel flow rate has a much more 
pronounced effect on the level of 
radiation from microgravity flames 
compared to those in normal gravity- 
Radiance varies almost linearly with 
flow rate under both microgravity and 
normal-gravity conditions, but radiation 
in microgravity can be up to 10 times 
larger than in normal gravity 

In addition to larger flame size and 
extensive soot formation, CO2 and 
H 2 appear to contribute dramatically 
to radiative loss from microgravity 
flame due to the accumulation and 
slow transport of these products 

The data suggests that radiative 
ignition of nearby materials may be 
promoted in low-gravity environments 
due to the increased radiative loss 
from fires 

SAIC. 



Temperature Measurement - Normal-Gravity Tests 












































































-f^ 


— 










J — — - 








■ ,<M 





t <l«) 



1IM 




































































































~r^ 



























t (stc) 



Temperature neaunncnts for 1 normal- gravity propane-air flarac it ! atm. Nat 
thai regardlcu of preuure or oiyfcn conccfiutuon, thermocouple! b. c, t. f. h, 
and 1 do not ihow • temperature nie 1*1 normal graviiy. Thii flickering flame had 
on (average) height of 5.0 cm. and a majuroum diameter of 0.14 cm. reproduced 
frompOj Here, fuel-flow rue - 1.5 cniVi. and nozxic radius - 00823 cm. 



1M 






























































































A 








' 






f,.i 




2M 













t IlK) 



The middle- and outer-columns of thermocouples do not register any 
temperature rise due to the pencil-like character of the flame 

The three centerline thermocouples show almost the same temperature 
levels due to the presence of buoyancy 



L 



5MC. 



J 



I 



Temperature Measurement - Microgravity Tests 








































^ 






\ 






1 ">■ 










\ 


/ v 










\ 


/ 






b 




\ 


4 — ■ 










-^*- 


i — i 























































i. 




•>. 


-t— 








-H 


-^= 



Temperature menurementl for t (5 liter.) microirjviry proewe >ir time li I 
urn Thermocouple h w" nol funaiomni properly dunn| thii drop, reproduced 
from 1 101 The irutuu 1 overlJioot u due 10 U>e ifi'Don procesi. i.e.. due u> ihe 
rcluie of fuel pnor 10 die novum of ihe ipirt itmtot Here, fuel flow me • I 3 
cm J /j. uid nozzle ndius * 0.0123 en. 



"■w^ y iv 



m 



t <mc) 



Data show the existence of taller and wider flames in microgravity 
Steep temperature gradients exist in the axial direction 
Far field shows temperature rise 

5 seconds of microgravity is not sufficient to obtain information on 
steady-state behavior, since the field is continuously changing due to the 
release and slow transport of combustion products 

SAIC. 






Overriding Observation 



Quantitative measurements show that steady-state has not 
been achieved in the ground-based experiments 

• The applicability of the previously cited observations 
hinges on this issue 

• Results to-date have clearly demonstrated the 
significance of quantitative data which can be 
augmented with the time available in a space 
experiment 



I 



SAIC. 



Theoretical Analyses 



V SAIC 



Theoretical Methods for Gas-Jet Diffusion Flames 

To-date 

• Model for filling process in evacuated bottles (for sampling) 

• G-jitter analysis for acceptable g-levels 

• Transient, elliptic analytical model for flame development 

• Parabolic, steady-state numerical model 

• Radiation from C0 2 and H,0 

• Inverted-flame (negative-g) studies 

• Detailed transport processes 

• Multicomponent diffusion 

In progress 

• Parabolic, steady-state numerical model 
Detailed radiation (absorption/emission) model 
Quasiglobal kinetics 
Soot processes and thermophoresis 
Radiation from soot 
Detailed kinetics 

Planned 

• Elliptic, 2-D, steady-state numerical model 

• Submodels developed from application of parabolic model 

• Transient, 2-D, Navier-Stokes model 

• Analytical modeling 

J5H7E 



r 



r 



Predictions - Theoretical and Experimental Flame Heights 









■ 


u» 


l.i. 


'••M Its J 


A 


: - 


£,*% l t.Bll 

£?•. j •..*» 


■ 


a ,'' 

• 


* jX"^ 



)U 1M MB It* «• 



I 







• 


w 


1.1. 


r M i 




• 




on 


I.H1 

• ail 


: AX 

■ * / 






MjiwUi Matter 



Ul 1M U» 11* **B 



SMC 



Sc—»*a»M !■■'""'** 



Transient Analytical Model 




», ■ 0.186 c«, U ■ 25.76 c«i/itc 

---- Eiperlaent (Cochran and Haslca, 1970) 

— - Transient, tltlptlc analytical aodel 

■-— Theoretical steady ititt 
■ ■ < .... i .. ^ 

"oT? — 



D« binary diffusion coefflcent 
of methane In nitrogen. 

The values of D-1.0 and 0.85 
aj|2/sec correspond to W50 
ond 650k, resoectlvely. 



2.0 



J.S 




1.0 I.S 

T1«e. sec 

• 0.00 ch'/iki ».. • «.1« 



TIT 



1.00-7 



7 -"1 -7- 



3.JS 



i. U ■ Z91.Z a/sec 

. Eiperlaent (Cochran, 1172) 

Transient, elliptic analytical node I 

— Theoretical steady state 

- ■ - - ■ - - ■ ■ ' _ 



0.0 



o.s 



1.0 



TI- 



LS 
sec 



2.0 



I.S 



Open Issues 

• Non-steady-state 

• Effects of P 

• Effects of 2 

• Fuel type 

• Analysis and theoretical methods 



■sac 



Saury Mfyt a mom immn m **-* Corvarm* v 




Program Objectives 

• Specific Objectives 

The questions thai need to be answered are: 

(a) What are the steady-state characteristics of the microgravity flame, and given 
enough time, do the weak flames extinguish by virtue of their size, radiative 
loss, and associated transport processes? 

(b) Why does the removal of buoyancy result in the underventilated-like 
behavior in some otherwise overventilated diffusion flames? 

(c) Why does microgravity flame length vary non-monotonically with pressure, 
and what are the roles of buoyancy and oxygen concentration in defining 
this behavior? 

(d) Why are the sooting characteristics sensitive to oxygen concentration in 
microgravity flames? 



Program Objectives- Specific Objective (a) 

What are the steady-state characteristics of the microgravity flame, and given 
enough time, do the weak flames extinguish by virtue of their size, radiative loss, and 
associated transport processes? 

• Continuous release and slow transport of products creates a transient field in the 
early stages of burning. 

• A quasi-equilibrium state can be reached for both temperature and species fields. 

• The flame approaches steady state, and the far-field will not have an effect on the 
near-field. 

• This quasi-steady-state persists for a period of time, and will then be followed by 
a transient approach toward extinction due to depletion of oxygen. 

• Which one of the following processes occurs in longer test times? 

• Radiative loss can not compete with the heat release, leading to bright, sooty 
flames. 

• Radiative loss exceeds the heat input, leading to extinction. 

Originally blue, Q^ < Q^^.. Brighter, sootier flame, 

weak flame ^\~ ultimately reaching steady state 

^^"Qiad. > Q nhun s,ow «xtinction process 



Program Objectives- Specific Objective (b) 

Why does the removal of buoyancy result in underventilated-like behavior in some 
otherwise overventilated diffusion flames? 

• Removal of buoyancy increases the residence time for soot formation. 

• Fuel pyrolysis is therefore augmented, leading to increased soot nuclei and growth. 

• Larger soot and more soot formation result in slower oxidation and persistence of 
the soot. 

• Soot is formed within the flame and unlike gas-phase species, virtually does not 
diffuse. 

• Thermophoresis, caused by temperature gradient, forces the soot toward the flame 
centerline. 

• As a result, the particles core up and stay within a column at the tip of the flame. 

• The tip opens, oxygen transport is impeded from reaching the center, and the 

flame is prevented from closing on the axis. 

• What happens to this core of soot in longer test times? 



stm 



Program Objectives - Specific Objectives (c) 

Why does microgravity flame length vary non-monotonlcally with pressure, and what 
are the roles of buoyancy and oxygen concentrations in defining this behavior? 

• Observed increase in length with pressure is due to increasing amounts of soot 
rather than a change in the gas-phase flame height. 

• At low-P, kinetic effects lengthen the gas-phase flame height. 

• As the pressure increases, reaction rate increases and flame length decreases. 

• As the pressure increases further, soot levels increase, and persist beyond 
the gas-phase flame front. 

• This gives an indication (through visualization) of an apparently longer flame. 

• Does this non-monotonic behavior tend to correct itself In longer test times, 
or is the effect aggravated? 



r 



Program Objectives - Specific Objectives (d) 

Why are sooting characteristics sensitive to oxygen concentration in 
microgravity flames? 

• Microgravity environment provides longer residence times. 

• This results in prolific sooting. 

• Higher 2 concentrations facilitate the soot oxidation, causing higher temperatures. 

• Lower oxygen concentrations produce cooler flames, reduced pyrolysis and soot 
formation, and reduced heat and mass transfer. 

• The competing effects of increased residence time with Lewis-number and 
Schmidt-number effects need to be clarified through longer test times. 



sue J 



r \ 

Program Objectives (Cont.) 

The overall and specific objectives of the program will be achieved through the 
following measurements and associated modeling: 

• Measurements (R: Required; D: Desired): 

• Temperature field (R) 

• Flame visualization (R) 

• Flame radiation (R) 

• Chamber pressure (R) 

• Acceleration (D) 

• Gas-phase species concentrations (D) 

• Infrared imaging for species (D) 

• Density-gradient field imaging (D) 

• Velocity field (D) 

• Soot size and distribution (D) 

• Soot volume fraction and number density (D) 

• Modeling that includes: 

• Multicomponent diffusion 

• Thermophores effects 

• Kinetics 

• Soot formation and particulate radiation 

• Axial diffusion effects 

• Transient effects 



V 



Justification for Space Experiment 

• Flame shape and color change continuously in 5 seconds of available time 

• Temperature data show a continuous increase in the 5 seconds of micgrogravity 

• Sooty flames show a continuous increase in radiation loss 

• Analytical predictions consistent with these observations 

• G-levels less than 10 5 are needed 

• The following (based on objectives of the program) need longer duration experiments 

• Steady-state characteristics 

• Weak-flame behavior 

• Underventilated-like phenomena 

• Non-monotonic flame-height variation with pressure 

• Effects of oxygen on the sensitivity of sooting characteristics 



-sac 




Experimental Options 



All objectives require: Temperature field, Flame visualization, Radiation, and Pressure measurements 



Obforttva 


Desired Measurements j 


Accel. 


Species 
lleld 


Infrared 
Imaging 


Density- 
gradient 
lleld 


Velocity 
lleld 


Sock sue 


-"i 


1 {•) SmMt «•«• cnatociomuc*. 
| ONm oneugn wnm, oo liw 
l woo* ftomoa •ittnguitn? 


X 


X 


X 


X 


X 


X 




X 


(b) Undorvontthiod-kliLO bonowtor 
D In otttonriM owfwonHlotoo' 
| ItaM duo to rmmma of 
| buoyancy. 


X 


X 


X 


X 


X 




X 


{C| MonmonotoMc bona* lor of 
rMorogrovity flamo-wjnftti wllh 
proiourr Ftoioo or buoyancy 
•»« 1L03 In ooflnlng m f 
bohavtor. 


X 


X 


X 


X 




X 


X 


(d) Sonsiitwiry of •oonng 
oftaroctortittct to Oiyfon 
In mkcrogrovlty flam**. 

1 


X 


X 


X 


X 


X 


X 


X 



V. 



Concept Design • Middeck 




imimmi Ms-orr wrnisiw ru«s 



JSMC 



>v 



Concept Design - CEM 



V. 




SAIC. 



Test Matrix for Space Experiment 



Operating Conditions 


No. of 
Experiments 


Note 


Fuel 


2 


Methane; propane 


Fuel-flow rate 


1 


Fixed mass-flow rate. When 
converted to atmospheric condition, 
1 .0 cm 3 /sec for propane and 2.0 
cm 3 /sec for methane 


Nozzle radius 


1 


Tapered-tip nozzle; inner radius - 
0.0825 cm 


Oxidizer 


3 


Oxygen/nitrogen mixtures; Og/Nj » 
0.15/0.85, 0.21/0.79, and 0.30/0.70 


Chamber pressure 


3 


P-0.5, 1.0, and 1.5 atm 


Total no. of flames 


18 





* Limited to one Reynolds number for each fuel. 



V. 



5AIC 



Soar* ^«tuNniMrawfCa>«aw* 



Diagnostics Requirements 



Variable 


Diagnostics 


Rationale/Applications to Model Development 


R*quirad 


DmM 


Fiane 

Structure 


Video camera 


Enables Ihe expenmenter to monitor the llame behavior 
during the experiment 


• 




Cinematography 
(using movie camera) 


Obtam time-resolved visualization data on ignition, flame 
development, color, luminosily, sooting, and quenching. 
Flame shapes can be compared directly with model 
predictions 


Schlieren camera 


To obtain qualitative temperature/density field 




• 


Temperature 


Array Of 

thermocouples 


Obtaining temperature distributions in the flow field They 
can be compared directly with the model predictions of the 
thermal liek). 


• 




Combustion 
and luel 
decomposition 
products 


Row of gas sampling 
probes (3) and array 
of probes {3 i 3) 


Obtaining species distributions which can be compared with 
the model predictions ol the species field. Unburned 
hydrocarbons are extremely important 




• 


Cham per 
pressure 


Pressure transducer 


Monitoring any increase in chamber pressure due to Ihe 
burning process, which yields estimates of the overall heat 
release rate: also used lor safety detection 


• 




Gas velocity 


Single -component 
laser-dopprer 
velocimetry (LOV), or 
panicle-image 
velocimetry (PIV) 


Centertine and radial velocities can be obtained using 
traversing mechanisms. LOV needs seeding. Compact 
lasers are becoming available. Velocity measurements will 
be compared directly with the model predictions of the Dow 

field 




• 



r 



SASC. 

Scar fcnkCJWrvMl 



Diagnostics Requirements (Cont.) 



Variable 


Diagnostic* 


Ratlonale/AppHcations to Model Development 


■HnHK 


Darin* 


Flame 
radiation 


Radiometry 


Isolation of thermal radiation, and determination ol the eflect 
ol radiation on reduction m temperature within the flame. 
Potential exists for determination of soot concentration. 
Predictions can be directly compared with measurements 


• 




Gravitational 
acceleration 


Aceeteromater 


Measurements of acceleration enables us to interpret the 
data that could be affected by low gravitational levels 




• 


Soeci'ic 
species 


Light-intensified 
array camera 


To identify species using appropriate optical fitters; 
particularly uselul lor H 3 0, C0 2 . and CO fields in 
non-sooting (blue) flames, as well as sooty flames 




• 


Soot number 
density and 
size 


Light absorption/ 
extinction 


A convenient, non-intrusive method to inter soot number 
density: when combined with radiation and species 
measurements, provides information on soot behavior and 
characteristics 




• 


Tnerrnophofetjc 
sampling 


Soot can be collected and size distribution can be obtained. 
Collected soot needs to be removed from the chamber after 
Shuttle landing (in middeck), and alter test is done (in 
Spacelab) 




• 



jSA/C 



REPORT DOCUMENTATION PAGE 



Form Approved 
OMB No. 0704-0188 



mm^^^^m^!^m^§^^^^ 



1. AGENCY USE ONLY {Leave blank) 



2. REPORT DATE 

August 1992 



3. REPORT TYPE AND DATES COVERED 

Final Contractor Report 



4. TITLE AND SUBTITLE 



Effects of Buoyancy on Gas Jet Diffusion Flames 



6. AUTHOR(S) 

M. Yousef Bahadori and Raymond B. Edelman 



7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 

Science Applications International Corporation 
Thermal Sciences Division 
21151 Western Avenue 
Torrance, California 90501-1724 



5. FUNDING NUMBERS 



WU-694-03-0A 

C-NAS3-22822 



8. PERFORMING ORGANIZATION 
REPORT NUMBER 



E-7720 



9. SPONSORING/MONITORING AGENCY NAMES(S) AND ADDRESS(ES) 

National Aeronautics and Space Administration 
Lewis Research Center 
Cleveland, Ohio 44135-3191 



10. SPONSORING/MONITORING 
AGENCY REPORT NUMBER 



NASA CR-191 109 



11. SUPPLEMENTARY NOTES 

Project Manager, D. Stocker, Space Experiments Division, (216) 433-2166. 



12a. DISTRIBUTION/AVAILABILITY STATEMENT 

Unclassified -Unlimited 
Subject Category 29 



12b. DISTRIBUTION CODE 



13. ABSTRACT (Maximum 200 words) 

The objective of this effort was to gain a better understanding of the fundamental phenomena involved in laminar 
gas jet diffusion flames in the absence of buoyancy by studying the transient phenomena of ignition and flame 
development, (quasi-) steady-state flame characteristics, soot effects, radiation, and, if any, extinction phenomena. 
This involved measurements of flame size and development, as well as temperature and radiation. Additionally, 
flame behavior, color and luminosity were observed and recorded. The tests quantified the effects of Reynolds 
number, nozzle size, fuel reactivity and type, oxygen concentration, and pressure on flame characteristics. Analyti- 
cal and numerical modeling efforts were also performed. Methane and propane flames were studied in the 2.2- 
Second Drop Tower and the 5.18-Second Zero-Gravity Facility of the NASA Lewis Research Center. In addition, a 
preliminary series of tests were conducted in the KC-135 research aircraft. Both microgravity and normal-gravity 
flames were studied in this program. The results have provided unique and new information on the behavior and 
characteristics of gas jet diffusion flames in microgravity environments. 



14. SUBJECT TERMS 

Microgravity combustion; Laminar flames; Diffusion flames 



17. SECURITY CLASSIFICATION 
OF REPORT 

Unclassified 



18. SECURITY CLASSIFICATION 
OF THIS PAGE 

Unclassified 



19. SECURITY CLASSIFICATION 
OF ABSTRACT 

Unclassified 



15. NUMBER OF PAGES 

200 



16. PRICE CODE 

A09 



20. LIMITATION OF ABSTRACT 



NSN 7540-01-280-5500 



Standard Form 298 (Rev. 2-89) 
Prescribed by ANSI Std. Z39-18 
298-102 







> 

o 

r- 

-i 
■< 

■n 


Q 00 

ST* 




30 

2 




ft 


Z 
O 


r) ~ c 




f 
3D 

m 

< 

• 







m 


• 


-< 


Doc 




3D 

m 
—1 




cr > 

3D H 

z m 


s 




m 
1 


3 








§5 ft 




H —1 

m 


a 




to — 

1 

1 






Z «3 
> 30 






30 z 


&- 




m 

v_^ do 

-< 








1 




30 
m 




C > 

3D -H 
z m 
m 




SL 







1 ° 
1 z ° 




<= > 

-\ -\ 
m 


1 <= 

m m 
3D z 






n; 




CO — 


z. 




1 

1 


s> 




z 


V 




3D Z 


i 




m 

-< 


•3 




30 


^r 




H O 






<= > 




„-l 


3D H 

z m 

]_ 1 | 5 





Z CC 
UJ LLl 

Si 



j 

o 

fa 

S3 



UJ z 
< => 



UJ 

z oc 
o z 

X C3 

I 

— CO 
3-' 



< 2 
o o 



uj z 
I- oc 
< =3 



>- 

CO *->. 
UJ 

z oe 

^S 
ec < 

o z 

X o 

1 

— CO 

3 w 



< 3 
Q O 



UJ Z 
I- CC 

< => 
O H 



>- 

to .— » 

UJ 

Z o£ 

? s 

en < 
a z 
i o 

l 

— CO 



< 3 

o o 



CO 

vO 



O 
UJ 

o 
> 

UJ 

ec 

5 

* 

o 



CtL 
O 



Jl